Lifelong user behavior sequences are crucial for capturing user interests and predicting user responses in modern recommendation systems. A two-stage paradigm is typically adopted to handle these long sequences: a subset of relevant behaviors is first searched from the original long sequences via an attention mechanism in the first stage and then aggregated with the target item to construct a discriminative representation for prediction in the second stage. In this work, we identify and characterize, for the first time, a neglected deficiency in existing long-sequence recommendation models: a single set of embeddings struggles with learning both attention and representation, leading to interference between these two processes. Initial attempts to address this issue with some common methods (e.g., linear projections---a technique borrowed from language processing) proved ineffective, shedding light on the unique challenges of recommendation models. To overcome this, we propose the Decoupled Attention and Representation Embeddings (DARE) model, where two distinct embedding tables are initialized and learned separately to fully decouple attention and representation. Extensive experiments and analysis demonstrate that DARE provides more accurate searches of correlated behaviors and outperforms baselines with AUC gains up to 9‰ on public datasets and notable improvements on Tencent's advertising platform. Furthermore, decoupling embedding spaces allows us to reduce the attention embedding dimension and accelerate the search procedure by 50\% without significant performance impact, enabling more efficient, high-performance online serving. Code in PyTorch for experiments, including model analysis, is available at https://github.com/thuml/DARE.

Diffusion models have emerged as powerful generative frameworks by progressively adding noise to data through a forward process and then reversing this process to generate realistic samples. While these models have achieved strong performance across various tasks and modalities, their application to temporal predictive learning remains underexplored. Existing approaches treat predictive learning as a conditional generation problem, but often fail to fully exploit the temporal dynamics inherent in the data, leading to challenges in generating temporally coherent sequences. To address this, we introduce Dynamical Diffusion (DyDiff), a theoretically sound framework that incorporates temporally aware forward and reverse processes. Dynamical Diffusion explicitly models temporal transitions at each diffusion step, establishing dependencies on preceding states to better capture temporal dynamics. Through the reparameterization trick, Dynamical Diffusion achieves efficient training and inference similar to any standard diffusion model. Extensive experiments across scientific spatiotemporal forecasting, video prediction, and time series forecasting demonstrate that Dynamical Diffusion consistently improves performance in temporal predictive tasks, filling a crucial gap in existing methodologies. Code is available at this repository: https://github.com/thuml/dynamical-diffusion.

We present Timer-XL, a causal Transformer for unified time series forecasting. To uniformly predict multidimensional time series, we generalize next token prediction, predominantly adopted for 1D token sequences, to multivariate next token prediction. The paradigm formulates various forecasting tasks as a long-context prediction problem. We opt for decoder-only Transformers that capture causal dependencies from varying-length contexts for unified forecasting, making predictions on non-stationary univariate time series, multivariate series with complicated dynamics and correlations, as well as covariate-informed contexts that include exogenous variables. Technically, we propose a universal TimeAttention to capture fine-grained intra- and inter-series dependencies of flattened time series tokens (patches), which is further enhanced by deft position embedding for temporal causality and variable equivalence. Timer-XL achieves state-of-the-art performance across task-specific forecasting benchmarks through a unified approach. Based on large-scale pre-training, Timer-XL achieves state-of-the-art zero-shot performance, making it a promising architecture for pre-trained time series models. Code is available at this repository: https://github.com/thuml/Timer-XL.

Recent advancements in diffusion models have revolutionized generative modeling. However, the impressive and vivid outputs they produce often come at the cost of significant model scaling and increased computational demands. Consequently, building personalized diffusion models based on off-the-shelf models has emerged as an appealing alternative. In this paper, we introduce a novel perspective on conditional generation for transferring a pre-trained model. From this viewpoint, we propose *Domain Guidance*, a straightforward transfer approach that leverages pre-trained knowledge to guide the sampling process toward the target domain. Domain Guidance shares a formulation similar to advanced classifier-free guidance, facilitating better domain alignment and higher-quality generations. We provide both empirical and theoretical analyses of the mechanisms behind Domain Guidance. Our experimental results demonstrate its substantial effectiveness across various transfer benchmarks, achieving over a 19.6\% improvement in FID and a 23.4\% improvement in FD$_\text{DINOv2}$ compared to standard fine-tuning. Notably, existing fine-tuned models can seamlessly integrate Domain Guidance to leverage these benefits, without additional training. Code is available at this repository: https://github.com/thuml/DomainGuidance.

We introduce Sundial, a family of native, flexible, and scalable time series foundation models. To predict the next-patch's distribution, we propose a TimeFlow Loss based on flow-matching, which facilitates native pre-training of Transformers on continuous-valued time series without discrete tokenization. Conditioned on arbitrary-length time series, our models are pre-trained without specifying any prior distribution and can generate multiple probable predictions, achieving more flexibility in representation learning than using parametric densities. Towards time series foundation models, we leverage minimal but crucial adaptations of Transformers and curate TimeBench with one trillion time points, comprising mostly real-world datasets and synthetic data. By mitigating mode collapse via TimeFlow Loss, we pre-train a family of Sundial models on TimeBench, which achieve unprecedented model capacity and generalization performance. In addition to excellent scalability, Sundial achieves state-of-the-art results on both point and probabilistic forecasting benchmarks with a just-in-time inference speed, i.e., making zero-shot predictions within a few milliseconds. We believe that Sundial's pioneering generative forecasting capability can improve model reliability in real-world decision-making. Code is available at: https://github.com/thuml/Sundial.

Although deep models have been widely explored in solving partial differential equations (PDEs), previous works are primarily limited to data only with up to tens of thousands of mesh points, far from the million-point scale required by industrial simulations that involve complex geometries. In the spirit of advancing neural PDE solvers to real industrial applications, we present Transolver++, a highly parallel and efficient neural solver that can accurately solve PDEs on million-scale geometries. Building upon previous advancements in solving PDEs by learning physical states via Transolver, Transolver++ is further equipped with an extremely optimized parallelism framework and a local adaptive mechanism to efficiently capture eidetic physical states from massive mesh points, successfully tackling the thorny challenges in computation and physics learning when scaling up input mesh size. Transolver++ increases the single-GPU input capacity to million-scale points for the first time and is capable of continuously scaling input size in linear complexity by increasing GPUs. Experimentally, Transolver++ yields 13% relative promotion across six standard PDE benchmarks and achieves over 20% performance gain in million-scale high-fidelity industrial simulations, whose sizes are 100$\times$ larger than previous benchmarks, covering car and 3D aircraft designs.

Heterogeneity in sensors and actuators across environments poses a significant challenge to building large-scale pre-trained world models on top of this low-dimensional sensor information. In this work, we explore pre-training world models for heterogeneous environments by addressing key transfer barriers in both data diversity and model flexibility. We introduce UniTraj, a unified dataset comprising over one million trajectories from 80 environments, designed to scale data while preserving critical diversity. Additionally, we propose TrajWorld, a novel architecture capable of flexibly handling varying sensor and actuator information and capturing environment dynamics in-context. Pre-training TrajWorld on UniTraj yields substantial gains in transition prediction, achieves a new state-of-the-art for off-policy evaluation, and also delivers superior online performance of model predictive control. To the best of our knowledge, this work, for the first time, demonstrates the transfer benefits of world models across heterogeneous and complex control environments. Code and data are available at https://github.com/thuml/TrajWorld.

Deep models have recently emerged as promising tools to solve partial differential equations (PDEs), known as neural PDE solvers. While neural solvers trained from either simulation data or physics-informed loss can solve PDEs reasonably well, they are mainly restricted to a few instances of PDEs, e.g. a certain equation with a limited set of coefficients. This limits their generalization to diverse PDEs, preventing them from being practical surrogate models of numerical solvers. In this paper, we present Unisolver, a novel Transformer model trained on diverse data and conditioned on diverse PDEs, aiming towards a universal neural PDE solver capable of solving a wide scope of PDEs. Instead of purely scaling up data and parameters, Unisolver stems from the theoretical analysis of the PDE-solving process. Inspired by the mathematical structure of PDEs that a PDE solution is fundamentally governed by a series of PDE components such as equation symbols and boundary conditions, we define a complete set of PDE components and flexibly embed them as domain-wise and point-wise deep conditions for Transformer PDE solvers. Integrating physical insights with recent Transformer advances, Unisolver achieves consistent state-of-the-art on three challenging large-scale benchmarks, showing impressive performance and generalizability. Code is available at [https://github.com/thuml/Unisolver](https://github.com/thuml/Unisolver).

Effective code optimization in compilers is crucial for computer and software engineering. The success of these optimizations primarily depends on the selection and ordering of the optimization passes applied to the code. While most compilers rely on fixed pass sequences, current methods to find the optimal sequence for specific programs either employ impractically slow search algorithms or learning methods that struggle to generalize to code unseen during training. To address these challenges, we introduce CompilerDream, the first world-model-based approach for general code optimization. CompilerDream features a compiler world model with a reward smoothing technique, enabling accurate simulation of optimization processes. Built on this model, code optimization agents can then be constructed via value prediction or direct optimization sequence generation. Trained on a large-scale program dataset, these agents serve as versatile code optimizers across diverse application scenarios and source-code languages. Our extensive experiments highlight CompilerDream's strong optimization capabilities for autotuning, where it leads the CompilerGym leaderboard. More importantly, the zero-shot generalization ability of large-scale trained compiler world model and agent, excels across diverse datasets, surpassing LLVM's built-in optimizations and state-of-the-art methods in both settings of value prediction and end-to-end code optimization.

Physics sensing plays a central role in many scientific and engineering domains, which inherently involves two coupled tasks: reconstructing dense physical fields from sparse observations and optimizing scattered sensor placements to observe maximum information. While deep learning has made rapid advances in sparse-data reconstruction, existing methods generally omit optimization of sensor placements, leaving the mutual enhancement between reconstruction and placement on the shelf. To change this suboptimal practice, we propose PhySense, a synergistic two-stage framework that learns to jointly reconstruct physical fields and to optimize sensor placements, both aiming for accurate physics sensing. The first stage involves a flow-based generative model enhanced by cross-attention to adaptively fuse sparse observations. Leveraging the reconstruction feedback, the second stage performs sensor placement via projected gradient descent to satisfy spatial constraints. We further prove that the learning objectives of the two stages are consistent with classical variance-minimization principles, providing theoretical guarantees. Extensive experiments across three challenging benchmarks, especially a 3D geometry dataset, indicate PhySense achieves state-of-the-art physics sensing accuracy and discovers informative sensor placements previously unconsidered. Code is available at this repository: https://github.com/thuml/PhySense.

Attention with bias, which extends standard attention by introducing prior knowledge as an additive bias matrix to the query-key scores, has been widely deployed in vision, language, protein-folding and other advanced scientific models, underscoring its status as a key evolution of this foundational module. However, introducing bias terms creates a severe efficiency bottleneck in attention computation. It disrupts the tightly fused memory-compute pipeline that underlies the speed of accelerators like FlashAttention, thereby stripping away most of their performance gains and leaving biased attention computationally expensive. Surprisingly, despite its common usage, targeted efficiency optimization for attention with bias remains absent, which seriously hinders its application in complex tasks. Diving into the computation of FlashAttention, we prove that its optimal efficiency is determined by the rank of the attention weight matrix. Inspired by this theoretical result, this paper presents FlashBias based on the low-rank compressed sensing theory, which can provide fast-exact computation for many widely used attention biases and a fast-accurate approximation for biases in general formalizations. FlashBias can fully take advantage of the extremely optimized matrix multiplication operation in modern GPUs, achieving 1.5$\times$ speedup for Pairformer in AlphaFold 3, and over 2$\times$ speedup for attention with bias in vision and language models without loss of accuracy. Code is available at this repository: https://github.com/thuml/FlashBias.

World models predict state transitions in response to actions and are increasingly developed across diverse modalities. However, standard training objectives such as maximum likelihood estimation (MLE) often misalign with task-specific goals of world models, i.e., transition prediction metrics like accuracy or perceptual quality. In this paper, we present RLVR-World, a unified framework that leverages reinforcement learning with verifiable rewards (RLVR) to directly optimize world models for such metrics. Despite formulating world modeling as autoregressive prediction of tokenized sequences, RLVR-World evaluates metrics of decoded predictions as verifiable rewards. We demonstrate substantial performance gains on both language- and video-based world models across domains, including text games, web navigation, and robot manipulation. Our work indicates that, beyond recent advances in reasoning language models, RLVR offers a promising post-training paradigm for enhancing the utility of generative models more broadly. Code, datasets, models, and video samples are available at the project website: https://thuml.github.io/RLVR-World.

PyTorch 2.x introduces a compiler designed to accelerate deep learning programs. However, for machine learning researchers, fully leveraging the PyTorch compiler can be challenging due to its operation at the Python bytecode level, making it appear as an opaque box. To address this, we introduce depyf, a tool designed to demystify the inner workings of the PyTorch compiler. depyf decompiles the bytecode generated by PyTorch back into equivalent source code and establishes connections between the code objects in the memory and their counterparts in source code format on the disk. This feature enables users to step through the source code line by line using debuggers, thus enhancing their understanding of the underlying processes. Notably, depyf is non-intrusive and user-friendly, primarily relying on two convenient context managers for its core functionality. The project is openly available at https://github.com/thuml/depyf and is recognized as a PyTorch ecosystem project at https://pytorch.org/blog/introducing-depyf.

The recent boom of linear forecasting models questions the ongoing passion for architectural modifications of Transformer-based forecasters. These forecasters leverage Transformers to model the global dependencies over temporal tokens of time series, with each token formed by multiple variates of the same timestamp. However, Transformers are challenged in forecasting series with larger lookback windows due to performance degradation and computation explosion. Besides, the embedding for each temporal token fuses multiple variates that represent potential delayed events and distinct physical measurements, which may fail in learning variate-centric representations and result in meaningless attention maps. In this work, we reflect on the competent duties of Transformer components and repurpose the Transformer architecture without any modification to the basic components. We propose iTransformer that simply applies the attention and feed-forward network on the inverted dimensions. Specifically, the time points of individual series are embedded into variate tokens which are utilized by the attention mechanism to capture multivariate correlations; meanwhile, the feed-forward network is applied for each variate token to learn nonlinear representations. The iTransformer model achieves state-of-the-art on challenging real-world datasets, which further empowers the Transformer family with promoted performance, generalization ability across different variates, and better utilization of arbitrary lookback windows, making it a nice alternative as the fundamental backbone of time series forecasting. Code is available at this repository: https://github.com/thuml/iTransformer.

Convolution-BatchNorm (ConvBN) blocks are integral components in various computer vision tasks and other domains. A ConvBN block can operate in three modes: Train, Eval, and Deploy. While the Train mode is indispensable for training models from scratch, the Eval mode is suitable for transfer learning and beyond, and the Deploy mode is designed for the deployment of models. This paper focuses on the trade-off between stability and efficiency in ConvBN blocks: Deploy mode is efficient but suffers from training instability; Eval mode is widely used in transfer learning but lacks efficiency. To solve the dilemma, we theoretically reveal the reason behind the diminished training stability observed in the Deploy mode. Subsequently, we propose a novel Tune mode to bridge the gap between Eval mode and Deploy mode. The proposed Tune mode is as stable as Eval mode for transfer learning, and its computational efficiency closely matches that of the Deploy mode. Through extensive experiments in object detection, classification, and adversarial example generation across $5$ datasets and $12$ model architectures, we demonstrate that the proposed Tune mode retains the performance while significantly reducing GPU memory footprint and training time, thereby contributing efficient ConvBN blocks for transfer learning and beyond. Our method has been integrated into both PyTorch (general machine learning framework) and MMCV/MMEngine (computer vision framework). Practitioners just need one line of code to enjoy our efficient ConvBN blocks thanks to PyTorch's builtin machine learning compilers.

Predictive Coding (PC) is a theoretical framework in cognitive science suggesting that the human brain processes cognition through spatiotemporal prediction of visual world. Existing studies have developed spatiotemporal prediction neural networks based on the PC theroy, emulating its two core mechanisms: Correcting predictions from residuals and Hierarchical learning. However, these models do not show the enhancement of prediction skills on real-world forecasting tasks, and ignore the Precision Weighting mechanism of PC theory. Precision weight posits that the brain allocates more attention to signals with lower Precision, contributing to the the cognitive ability of human brains. This work introduces the Cognitive Diffusion Probabilistic Models (CogDPM) which demonstrates the connection between diffusion probabilistic models and PC theory. CogDPM features a precision estimation method based on the hierarchical sampling capabilities of diffusion models, and allocate the guidance with precision weights estimated by the inherent property of diffusion models. We experimentally show that the precision weights is an estimator of model’s predictability on the rigid body and fluid motion dataset. We also apply CogDPM to real-world prediction tasks using the U.K. precipitation and ERA surface wind datasets. Our results demonstrate that CogDPM outperforms both existing domain-specific operational models and general deep prediction models in providing more proficient forecasting.

Time series pre-training has recently garnered wide attention for its potential to reduce labeling expenses and benefit various downstream tasks. Prior methods are mainly based on pre-training techniques well-acknowledged in vision or language, such as masked modeling and contrastive learning. However, randomly masking time series or calculating series-wise similarity will distort or neglect inherent temporal correlations crucial in time series data. To emphasize temporal correlation modeling, this paper proposes TimeSiam as a simple but effective self-supervised pre-training framework for Time series based on Siamese networks. Concretely, TimeSiam pre-trains Siamese encoders to capture intrinsic temporal correlations between randomly sampled past and current subseries. With a simple data augmentation method (e.g. masking), TimeSiam can benefit from diverse augmented subseries and learn internal time-dependent representations through a past-to-current reconstruction. Moreover, learnable lineage embeddings are also introduced to distinguish temporal distance between sampled series and further foster the learning of diverse temporal correlations. TimeSiam consistently outperforms extensive advanced pre-training baselines, demonstrating superior forecasting and classification capabilities across 13 standard benchmarks in both intra- and cross-domain scenarios.

Recent advances in foundation models have led to a promising trend of developing large recommendation models to leverage vast amounts of available data. Still, mainstream models remain embarrassingly small in size and naive enlarging does not lead to sufficient performance gain, suggesting a deficiency in the model scalability. In this paper, we identify the embedding collapse phenomenon as the inhibition of scalability, wherein the embedding matrix tends to occupy a low-dimensional subspace. Through empirical and theoretical analysis, we demonstrate a two-sided effect of feature interaction specific to recommendation models. On the one hand, interacting with collapsed embeddings restricts embedding learning and exacerbates the collapse issue. On the other hand, interaction is crucial in mitigating the fitting of spurious features as a scalability guarantee. Based on our analysis, we propose a simple yet effective multi-embedding design incorporating embedding-set-specific interaction modules to learn embedding sets with large diversity and thus reduce collapse. Extensive experiments demonstrate that this proposed design provides consistent scalability and effective collapse mitigation for various recommendation models. Code is available at this repository: https://github.com/thuml/Multi-Embedding.

Deep learning has contributed remarkably to the advancement of time series analysis. Still, deep models can encounter performance bottlenecks in real-world data-scarce scenarios, which can be concealed due to the performance saturation with small models on current benchmarks. Meanwhile, large models have demonstrated great powers in these scenarios through large-scale pre-training. Continuous progress has been achieved with the emergence of large language models, exhibiting unprecedented abilities such as few-shot generalization, scalability, and task generality, which are however absent in small deep models. To change the status quo of training scenario-specific small models from scratch, this paper aims at the early development of large time series models (LTSM). During pre-training, we curate large-scale datasets with up to 1 billion time points, unify heterogeneous time series into single-series sequence (S3) format, and develop the GPT-style architecture toward LTSMs. To meet diverse application needs, we convert forecasting, imputation, and anomaly detection of time series into a unified generative task. The outcome of this study is a Time Series Transformer (Timer), which is generative pre-trained by next token prediction and adapted to various downstream tasks with promising capabilities as an LTSM. Code and datasets are available at: https://github.com/thuml/Large-Time-Series-Model.

Model-based reinforcement learning (MBRL) holds the promise of sample-efficient learning by utilizing a world model, which models how the environment works and typically encompasses components for two tasks: observation modeling and reward modeling. In this paper, through a dedicated empirical investigation, we gain a deeper understanding of the role each task plays in world models and uncover the overlooked potential of sample-efficient MBRL by mitigating the domination of either observation or reward modeling. Our key insight is that while prevalent approaches of explicit MBRL attempt to restore abundant details of the environment via observation models, it is difficult due to the environment’s complexity and limited model capacity. On the other hand, reward models, while dominating implicit MBRL and adept at learning compact task-centric dynamics, are inadequate for sample-efficient learning without richer learning signals. Motivated by these insights and discoveries, we propose a simple yet effective approach, HarmonyDream, which automatically adjusts loss coefficients to maintain task harmonization, i.e. a dynamic equilibrium between the two tasks in world model learning. Our experiments show that the base MBRL method equipped with HarmonyDream gains 10%-69% absolute performance boosts on visual robotic tasks and sets a new state-of-the-art result on the Atari 100K benchmark. Code is available at https://github.com/thuml/HarmonyDream.

Transformers have empowered many milestones across various fields and have recently been applied to solve partial differential equations (PDEs). However, since PDEs are typically discretized into large-scale meshes with complex geometries, it is challenging for Transformers to capture intricate physical correlations directly from massive individual points. Going beyond superficial and unwieldy meshes, we present Transolver based on a more foundational idea, which is learning intrinsic physical states hidden behind discretized geometries. Specifically, we propose a new Physics-Attention to adaptively split the discretized domain into a series of learnable slices of flexible shapes, where mesh points under similar physical states will be ascribed to the same slice. By calculating attention to physics-aware tokens encoded from slices, Transovler can effectively capture intricate physical correlations under complex geometrics, which also empowers the solver with endogenetic geometry-general modeling capacity and can be efficiently computed in linear complexity. Transolver achieves consistent state-of-the-art with 22% relative gain across six standard benchmarks and also excels in large-scale industrial simulations, including car and airfoil designs. Code is available at https://github.com/thuml/Transolver.

Fluid prediction is a long-standing challenge due to the intrinsic high-dimensional non-linear dynamics. Previous methods usually utilize the non-linear modeling capability of deep models to directly estimate velocity fields for future prediction. However, skipping over inherent physical properties but directly learning superficial velocity fields will overwhelm the model from generating precise or physics-reliable results. In this paper, we propose the HelmFluid toward an accurate and interpretable predictor for fluid. Inspired by the Helmholtz theorem, we design a HelmDynamics block to learn Helmholtz dynamics, which decomposes fluid dynamics into more solvable curl-free and divergence-free parts, physically corresponding to potential and stream functions of fluid. By embedding the HelmDynamics block into a Multiscale Multihead Integral Architecture, HelmFluid can integrate learned Helmholtz dynamics along temporal dimension in multiple spatial scales to yield future fluid. Compared with previous velocity estimating methods, HelmFluid is faithfully derived from Helmholtz theorem and ravels out complex fluid dynamics with physically interpretable evidence. Experimentally, HelmFluid achieves consistent state-of-the-art in both numerical simulated and real-world observed benchmarks, even for scenarios with complex boundaries.

Vision Transformers (ViTs) excel in computer vision tasks due to their ability to capture global context among tokens. However, their quadratic complexity $\mathcal{O}(N^2D)$ in terms of token number $N$ and feature dimension $D$ limits practical use on mobile devices, necessitating more mobile-friendly ViTs with reduced latency. Multi-head linear-attention is emerging as a promising alternative with linear complexity $\mathcal{O}(NDd)$, where $d$ is the per-head dimension. Still, more compute is needed as $d$ gets large for model accuracy. Reducing $d$ improves mobile friendliness at the expense of excessive small heads weak at learning valuable subspaces, ultimately impeding model capability. To overcome this efficiency-capability dilemma, we propose a novel Mobile-Attention design with a head-competition mechanism empowered by information flow, which prevents overemphasis on less important subspaces upon trivial heads while preserving essential subspaces to ensure Transformer's capability. It enables linear-time complexity on mobile devices by supporting a small per-head dimension $d$ for mobile efficiency. By replacing the standard attention of ViTs with Mobile-Attention, our optimized ViTs achieved enhanced model capacity and competitive performance in a range of computer vision tasks. Specifically, we have achieved remarkable reductions in latency on the iPhone 12. Code is available at https://github.com/thuml/MobileAttention.

Hyperparameter optimization (HPO) is known to be costly in deep learning, especially when leveraging automated approaches. Most of the existing automated HPO methods are accuracy-based, i.e., accuracy metrics are used to guide the trials of different hyperparameter configurations amongst a specific search space. However, many trials may encounter severe training problems, such as vanishing gradients and insufficient convergence, which can hardly be reflected by accuracy metrics in the early stages of the training and often result in poor performance. This leads to an inefficient optimization trajectory because the bad trials occupy considerable computation resources and reduce the probability of finding excellent hyperparameter configurations within a time limitation. In this paper, we propose Bad Trial Tackler (BTTackler), a novel HPO framework that introduces training diagnosis to identify training problems automatically and hence tackles bad trials. BTTackler diagnoses each trial by calculating a set of carefully designed quantified indicators and triggers early termination if any training problems are detected. Evaluations are performed on representative HPO tasks consisting of three classical deep neural networks (DNN) and four widely used HPO methods. To better quantify the effectiveness of an automated HPO method, we propose two new measurements based on accuracy and time consumption. Results show the advantage of BTTackler on two-fold: (1) it reduces 40.33% of time consumption to achieve the same accuracy comparable to baseline methods on average and (2) it conducts 44.5% more top-10 trials than baseline methods on average within a given time budget. We also released an open-source Python library that allows users to easily apply BTTackler to automated HPO processes with minimal code changes\footnotehttps://github.com/thuml/BTTackler.

World models empower model-based agents to interactively explore, reason, and plan within imagined environments for real-world decision-making. However, the high demand for interactivity poses challenges in harnessing recent advancements in video generative models for developing world models at scale. This work introduces Interactive VideoGPT (iVideoGPT), a scalable autoregressive transformer framework that integrates multimodal signals—visual observations, actions, and rewards—into a sequence of tokens, facilitating an interactive experience of agents via next-token prediction. iVideoGPT features a novel compressive tokenization technique that efficiently discretizes high-dimensional visual observations. Leveraging its scalable architecture, we are able to pre-train iVideoGPT on millions of human and robotic manipulation trajectories, establishing a versatile foundation that is adaptable to serve as interactive world models for a wide range of downstream tasks. These include action-conditioned video prediction, visual planning, and model-based reinforcement learning, where iVideoGPT achieves competitive performance compared with state-of-the-art methods. Our work advances the development of interactive general world models, bridging the gap between generative video models and practical model-based reinforcement learning applications. Code and pre-trained models are available at https://thuml.github.io/iVideoGPT.

Foundation models of time series have not been fully developed due to the limited availability of time series corpora and the underexploration of scalable pre-training. Based on the similar sequential formulation of time series and natural language, increasing research demonstrates the feasibility of leveraging large language models (LLM) for time series. Nevertheless, the inherent autoregressive property and decoder-only architecture of LLMs have not been fully considered, resulting in insufficient utilization of LLM abilities. To fully revitalize the general-purpose token transition and multi-step generation capability of large language models, we propose AutoTimes to repurpose LLMs as autoregressive time series forecasters, which projects time series into the embedding space of language tokens and autoregressively generates future predictions with arbitrary lengths. Compatible with any decoder-only LLMs, the consequent forecaster exhibits the flexibility of the lookback length and scalability with larger LLMs. Further, we formulate time series as prompts, extending the context for prediction beyond the lookback window, termed in-context forecasting. By introducing LLM-embedded textual timestamps, AutoTimes can utilize chronological information to align multivariate time series. Empirically, AutoTimes achieves state-of-the-art with 0.1% trainable parameters and over $5\times$ training/inference speedup compared to advanced LLM-based forecasters. Code is available at this repository: https://github.com/thuml/AutoTimes.

Accurately predicting the future fluid is vital to extensive areas such as meteorology, oceanology, and aerodynamics. However, since the fluid is usually observed from the Eulerian perspective, its moving and intricate dynamics are seriously obscured and confounded in static grids, bringing thorny challenges to the prediction. This paper introduces a new Lagrangian-Eulerian combined paradigm to tackle the tanglesome fluid dynamics. Instead of solely predicting the future based on Eulerian observations, we propose DeepLag to discover hidden Lagrangian dynamics within the fluid by tracking the movements of adaptively sampled key particles. Further, DeepLag presents a new paradigm for fluid prediction, where the Lagrangian movement of the tracked particles is inferred from Eulerian observations, and their accumulated Lagrangian dynamics information is incorporated into global Eulerian evolving features to guide future prediction respectively. Tracking key particles not only provides a transparent and interpretable clue for fluid dynamics but also makes our model free from modeling complex correlations among massive grids for better efficiency. Experimentally, DeepLag excels in three challenging fluid prediction tasks covering 2D and 3D, simulated and real-world fluids. Code is available at this repository: https://github.com/thuml/DeepLag.

Deep models have demonstrated remarkable performance in time series forecasting. However, due to the partially-observed nature of real-world applications, solely focusing on the target of interest, so-called endogenous variables, is usually insufficient to guarantee accurate forecasting. Notably, a system is often recorded into multiple variables, where the exogenous variables can provide valuable external information for endogenous variables. Thus, unlike well-established multivariate or univariate forecasting paradigms that either treat all the variables equally or ignore exogenous information, this paper focuses on a more practical setting: time series forecasting with exogenous variables. We propose a novel approach, TimeXer, to ingest external information to enhance the forecasting of endogenous variables. With deftly designed embedding layers, TimeXer empowers the canonical Transformer with the ability to reconcile endogenous and exogenous information, where patch-wise self-attention and variate-wise cross-attention are used simultaneously. Moreover, global endogenous tokens are learned to effectively bridge the causal information underlying exogenous series into endogenous temporal patches. Experimentally, TimeXer achieves consistent state-of-the-art performance on twelve real-world forecasting benchmarks and exhibits notable generality and scalability. Code is available at this repository: https://github.com/thuml/TimeXer.

Physics-informed neural networks (PINNs) have been widely applied to solve partial differential equations (PDEs) by enforcing outputs and gradients of deep models to satisfy target equations. Due to the limitation of numerical computation, PINNs are conventionally optimized on finite selected points. However, since PDEs are usually defined on continuous domains, solely optimizing models on scattered points may be insufficient to obtain an accurate solution for the whole domain. To mitigate this inherent deficiency of the default scatter-point optimization, this paper proposes and theoretically studies a new training paradigm as region optimization. Concretely, we propose to extend the optimization process of PINNs from isolated points to their continuous neighborhood regions, which can theoretically decrease the generalization error, especially for hidden high-order constraints of PDEs. A practical training algorithm, Region Optimized PINN (RoPINN), is seamlessly derived from this new paradigm, which is implemented by a straightforward but effective Monte Carlo sampling method. By calibrating the sampling process into trust regions, RoPINN finely balances optimization and generalization error. Experimentally, RoPINN consistently boosts the performance of diverse PINNs on a wide range of PDEs without extra backpropagation or gradient calculation. Code is available at this repository: https://github.com/thuml/RoPINN.

Diffusion models have significantly advanced the field of generative modeling. However, training a diffusion model is computationally expensive, creating a pressing need to adapt off-the-shelf diffusion models for downstream generation tasks. Current fine-tuning methods focus on parameter-efficient transfer learning but overlook the fundamental transfer characteristics of diffusion models. In this paper, we investigate the transferability of diffusion models and observe a monotonous chain of forgetting trend of transferability along the reverse process. Based on this observation and novel theoretical insights, we present Diff-Tuning, a frustratingly simple transfer approach that leverages the chain of forgetting tendency. Diff-Tuning encourages the fine-tuned model to retain the pre-trained knowledge at the end of the denoising chain close to the generated data while discarding the other noise side. We conduct comprehensive experiments to evaluate Diff-Tuning, including the transfer of pre-trained Diffusion Transformer models to eight downstream generations and the adaptation of Stable Diffusion to five control conditions with ControlNet. Diff-Tuning achieves a 24.6% improvement over standard fine-tuning and enhances the convergence speed of ControlNet by 24%. Notably, parameter-efficient transfer learning techniques for diffusion models can also benefit from Diff-Tuning. Code is available at this repository: https://github.com/thuml/Diffusion-Tuning.

Sequential recommendation requires the recommender to capture the evolving behavior characteristics from logged user behavior data for accurate recommendations. Nevertheless, user behavior sequences are viewed as a script with multiple ongoing threads intertwined. We find that only a small set of pivotal behaviors can be evolved into the user's future action. As a result, the future behavior of the user is hard to predict. We conclude this characteristic for sequential behaviors of each user as thebehavior pathway. Different users have their unique behavior pathways. Among existing sequential models, transformers have shown great capacity in capturing global-dependent characteristics. However, these models mainly provide a dense distribution over all previous behaviors using the self-attention mechanism, making the final predictions overwhelmed by the trivial behaviors not adjusted to each user. In this paper, we build the Recommender Transformer (RETR) with a novel Pathway Attention mechanism. RETR can dynamically plan the behavior pathway specified for each user, and sparingly activate the network through this behavior pathway to effectively capture evolving patterns useful for recommendation. The key design is a learned binary route to prevent the behavior pathway from being overwhelmed by trivial behaviors. Pathway attention is model-agnostic and can be applied to a series of transformer-based models for sequential recommendation. We empirically evaluate RETR on seven intra-domain benchmarks and RETR yields state-of-the-art performance. On another five cross-domain benchmarks, RETR can capture more domain-invariant representations for sequential recommendation.

In the open world, various label sets and domain configurations give rise to a variety of Domain Adaptation (DA) setups, including closed-set, partial-set, open-set, and universal DA, as well as multi-source and multi-target DA. It is notable that existing DA methods are generally designed only for a specific setup, and may under-perform in setups they are not tailored to. This paper shifts the common paradigm of DA to Versatile Domain Adaptation (VDA), where one method can handle several different DA setups without any modification. Towards this goal, we first delve into a general inductive bias: class confusion, and then uncover that reducing such pairwise class confusion leads to significant transfer gains. With this insight, we propose one general class confusion loss (CC-Loss) to learn many setups. We estimate class confusion based only on classifier predictions and minimize the class confusion to enable accurate target predictions. Further, we improve the loss by enforcing the consistency of confusion matrices under different data augmentations to encourage its invariance to distribution perturbations. Experiments on 2D vision and 3D vision benchmarks show that the CC-Loss performs competitively in different mainstream DA setups.

Undoubtedly, Deep Neural Networks (DNNs), from AlexNet to ResNet to Transformer, have sparked revolutionary advancements in diverse computer vision tasks. The scale of DNNs has grown exponentially due to the rapid development of computational resources. Despite the tremendous success, DNNs typically depend on massive amounts of training data (especially the recent various foundation models) to achieve high performance and are brittle in that their performance can degrade severely with small changes in their operating environment. Generally, collecting massive-scale training datasets is costly or even infeasible, as for certain fields, only very limited or no examples at all can be gathered. Nevertheless, collecting, labeling, and vetting massive amounts of practical training data is certainly difficult and expensive, as it requires the painstaking efforts of experienced human annotators or experts, and in many cases, prohibitively costly or impossible due to some reason, such as privacy, safety or ethic issues.

Estimating heterogeneous treatment effects (HTE) from observational studies is rising in importance due to the widespread accumulation of data in many fields. Due to the selection bias behind the inaccessibility of counterfactual data, the problem differs fundamentally from supervised learning in a challenging way. However, existing works on modeling selection bias and corresponding algorithms do not naturally generalize to non-binary treatment spaces. To address this limitation, we propose to use mutual information to describe selection bias in estimating HTE and derive a novel error bound using the mutual information between the covariates and the treatments, which is the first error bound to cover general treatment schemes including multinoulli or continuous spaces. We then bring forth theoretically justified algorithms, the Mutual Information Treatment Network (MitNet), using adversarial optimization to reduce selection bias and obtain more accurate HTE estimations. Our algorithm reaches remarkable performance in both simulation study and empirical evaluation.

Out-of-distribution (OOD) generalization, where the model needs to handle distribution shifts from training, is a major challenge of machine learning. Contrastive language-image pre-training (CLIP) models have shown impressive zero-shot ability, but the further adaptation of CLIP on downstream tasks undesirably degrades OOD performances. This paper aims at generalizing CLIP to out-of-distribution test data on downstream tasks. We propose CLIPood, a fine-tuning method that can adapt CLIP models to OOD situations where both domain shifts and open classes may occur on the unseen test data. To exploit the semantic relations between classes from the text modality, CLIPood introduces a new training objective, margin metric softmax (MMS), with class adaptive margins for fine-tuning. To incorporate both pre-trained zero-shot model and fine-tuned task-adaptive model, CLIPood leverages a new optimization strategy, Beta moving average (BMA), to maintain a temporal ensemble weighted by Beta distribution. Experiments on diverse datasets with different OOD scenarios show that CLIPood consistently outperforms existing generalization techniques.

Deep models have achieved impressive progress in solving partial differential equations (PDEs). A burgeoning paradigm is learning neural operators to approximate the input-output mappings of PDEs. While previous deep models have explored the multiscale architectures and various operator designs, they are limited to learning the operators as a whole in the coordinate space. In real physical science problems, PDEs are complex coupled equations with numerical solvers relying on discretization into high-dimensional coordinate space, which cannot be precisely approximated by a single operator nor efficiently learned due to the curse of dimensionality. We present Latent Spectral Models (LSM) toward an efficient and precise solver for high-dimensional PDEs. Going beyond the coordinate space, LSM enables an attention-based hierarchical projection network to reduce the high-dimensional data into a compact latent space in linear time. Inspired by classical spectral methods in numerical analysis, we design a neural spectral block to solve PDEs in the latent space that approximates complex input-output mappings via learning multiple basis operators, enjoying nice theoretical guarantees for convergence and approximation. Experimentally, LSM achieves consistent state-of-the-art and yields a relative gain of 11.5% averaged on seven benchmarks covering both solid and fluid physics. Code is available at https://github.com/thuml/Latent-Spectral-Models.

Unsupervised pre-training methods utilizing large and diverse datasets have achieved tremendous success across a range of domains. Recent work has investigated such unsupervised pre-training methods for model-based reinforcement learning (MBRL) but is limited to domain-specific or simulated data. In this paper, we study the problem of pre-training world models with abundant in-the-wild videos for efficient learning of downstream visual control tasks. However, in-the-wild videos are complicated with various contextual factors, such as intricate backgrounds and textured appearance, which precludes a world model from extracting shared world knowledge to generalize better. To tackle this issue, we introduce Contextualized World Models (ContextWM) that explicitly separate context and dynamics modeling to overcome the complexity and diversity of in-the-wild videos and facilitate knowledge transfer between distinct scenes. Specifically, a contextualized extension of the latent dynamics model is elaborately realized by incorporating a context encoder to retain contextual information and empower the image decoder, which encourages the latent dynamics model to concentrate on essential temporal variations. Our experiments show that in-the-wild video pre-training equipped with ContextWM can significantly improve the sample efficiency of MBRL in various domains, including robotic manipulation, locomotion, and autonomous driving. Code is available at this repository: https://github.com/thuml/ContextWM.

Time series analysis is widely used in extensive areas. Recently, to reduce labeling expenses and benefit various tasks, self-supervised pre-training has attracted immense interest. One mainstream paradigm is masked modeling, which successfully pre-trains deep models by learning to reconstruct the masked content based on the unmasked part. However, since the semantic information of time series is mainly contained in temporal variations, the standard way of randomly masking a portion of time points will seriously ruin vital temporal variations of time series, making the reconstruction task too difficult to guide representation learning. We thus present SimMTM, a Simple pre-training framework for Masked Time-series Modeling. By relating masked modeling to manifold learning, SimMTM proposes to recover masked time points by the weighted aggregation of multiple neighbors outside the manifold, which eases the reconstruction task by assembling ruined but complementary temporal variations from multiple masked series. SimMTM further learns to uncover the local structure of the manifold, which is helpful for masked modeling. Experimentally, SimMTM achieves state-of-the-art fine-tuning performance compared to the most advanced time series pre-training methods in two canonical time series analysis tasks: forecasting and classification, covering both in- and cross-domain settings.

Auxiliary-Task Learning (ATL) aims to improve the performance of the target task by leveraging the knowledge obtained from related tasks. Occasionally, learning multiple tasks simultaneously results in lower accuracy than learning only the target task, which is known as negative transfer. This problem is often attributed to the gradient conflicts among tasks, and is frequently tackled by coordinating the task gradients in previous works. However, these optimization-based methods largely overlook the auxiliary-target generalization capability. To better understand the root cause of negative transfer, we experimentally investigate it from both optimization and generalization perspectives. Based on our findings, we introduce ForkMerge, a novel approach that periodically forks the model into multiple branches, automatically searches the varying task weights by minimizing target validation errors, and dynamically merges all branches to filter out detrimental task-parameter updates. On a series of auxiliary-task learning benchmarks, ForkMerge outperforms existing methods and effectively mitigates negative transfer.

Real-world time series are characterized by intrinsic non-stationarity that poses a principal challenge for deep forecasting models. While previous models suffer from complicated series variations induced by changing temporal distribution, we tackle non-stationary time series with modern Koopman theory that fundamentally considers the underlying time-variant dynamics. Inspired by Koopman theory of portraying complex dynamical systems, we disentangle time-variant and time-invariant components from intricate non-stationary series by Fourier Filter and design Koopman Predictor to advance respective dynamics forward. Technically, we propose Koopa as a novel Koopman forecaster composed of stackable blocks that learn hierarchical dynamics. Koopa seeks measurement functions for Koopman embedding and utilizes Koopman operators as linear portraits of implicit transition. To cope with time-variant dynamics that exhibits strong locality, Koopa calculates context-aware operators in the temporal neighborhood and is able to utilize incoming ground truth to scale up forecast horizon. Besides, by integrating Koopman Predictors into deep residual structure, we ravel out the binding reconstruction loss in previous Koopman forecasters and achieve end-to-end forecasting objective optimization. Compared with the state-of-the-art model, Koopa achieves competitive performance while saving 77.3% training time and 76.0% memory.

Domain adaptation targets at knowledge acquisition and dissemination from a labeled source domain to an unlabeled target domain under distribution shift. Still, the common requirement of identical class space shared across domains hinders applications of domain adaptation to partial-set domains. Recent advances show that deep pre-trained models of large scale endow rich knowledge to tackle diverse downstream tasks of small scale. Thus, there is a strong incentive to adapt models from large-scale domains to small-scale domains. This paper introduces Partial Domain Adaptation (PDA), a learning paradigm that relaxes the identical class space assumption to that the source class space subsumes the target class space. First, we present a theoretical analysis of partial domain adaptation, which uncovers the importance of estimating the transferable probability of each class and each instance across domains. Then, we propose Selective Adversarial Network (SAN and SAN++) with a bi-level selection strategy and an adversarial adaptation mechanism. The bi-level selection strategy up-weighs each class and each instance simultaneously for source supervised training, target self-training, and source-target adversarial adaptation through the transferable probability estimated alternately by the model. Experiments on standard partial-set datasets and more challenging tasks with superclasses show that SAN++ outperforms several domain adaptation methods.

Few-shot learning aims to fast adapt a deep model from a few examples. While pre-training and meta-training can create deep models powerful for few-shot generalization, we find that pre-training and meta-training focus respectively on cross-domain transferability and cross-task transferability, which restricts their data efficiency in the entangled settings of domain shift and task shift. We thus propose the Omni-Training framework to seamlessly bridge pre-training and meta-training for data-efficient few-shot learning. Our first contribution is a tri-flow Omni-Net architecture. Besides the joint representation flow, Omni-Net introduces two parallel flows for pre-training and meta-training, responsible for improving domain transferability and task transferability respectively. Omni-Net further coordinates the parallel flows by routing their representations via the joint-flow, enabling knowledge transfer across flows. Our second contribution is the Omni-Loss, which introduces a self-distillation strategy separately on the pre-training and meta-training objectives for boosting knowledge transfer throughout different training stages. Omni-Training is a general framework to accommodate many existing algorithms. Evaluations justify that our single framework consistently and clearly outperforms the individual state-of-the-art methods on both cross-task and cross-domain settings in a variety of classification, regression and reinforcement learning problems.

The predictive learning of spatiotemporal sequences aims to generate future images by learning from the historical context, where the visual dynamics are believed to have modular structures that can be learned with compositional subsystems. This paper models these structures by presenting PredRNN, a new recurrent network, in which a pair of memory cells are explicitly decoupled, operate in nearly independent transition manners, and finally form unified representations of the complex environment. Concretely, besides the original memory cell of LSTM, this network is featured by a zigzag memory flow that propagates in both bottom-up and top-down directions across all layers, enabling the learned visual dynamics at different levels of RNNs to communicate. It also leverages a memory decoupling loss to keep the memory cells from learning redundant features. We further propose a new curriculum learning strategy to force PredRNN to learn long-term dynamics from context frames, which can be generalized to most sequence-to-sequence models. We provide detailed ablation studies to verify the effectiveness of each component. Our approach is shown to obtain highly competitive results on five datasets for both action-free and action-conditioned predictive learning scenarios.

Learning predictive models for unlabeled spatiotemporal data is challenging in part because visual dynamics can be highly entangled, especially in real scenes. In this paper, we refer to the multi-modal output distribution of predictive learning as spatiotemporal modes. We find an experimental phenomenon named spatiotemporal mode collapse (STMC) on most existing video prediction models, that is, features collapse into invalid representation subspaces due to the ambiguous understanding of mixed physical processes. We propose to quantify STMC and explore its solution for the first time in the context of unsupervised predictive learning. To this end, we present ModeRNN, a decoupling-aggregation framework that has a strong inductive bias of discovering the compositional structures of spatiotemporal modes between recurrent states. We first leverage a set of dynamic slots with independent parameters to extract individual building components of spatiotemporal modes. We then perform a weighted fusion of slot features to adaptively aggregate them into a unified hidden representation for recurrent updates. Through a series of experiments, we show high correlation between STMC and the fuzzy prediction results of future video frames. Besides, ModeRNN is shown to better mitigate STMC and achieve the state of the art on five video prediction datasets.

Predictive learning ideally builds the world model of physical processes in one or more given environments. Typical setups assume that we can collect data from all environments at all times. In practice, however, different prediction tasks may arrive sequentially so that the environments may change persistently throughout the training procedure. Can we develop predictive learning algorithms that can deal with more realistic, non-stationary physical environments? In this paper, we study a new continual learning problem in the context of video prediction, and observe that most existing methods suffer from severe catastrophic forgetting in this setup. To tackle this problem, we propose the continual predictive learning (CPL) approach, which learns a mixture world model via predictive experience replay and performs test-time adaptation with non-parametric task inference. We construct two new benchmarks based on RoboNet and KTH, in which different tasks correspond to different physical robotic environments or human actions. Our approach is shown to effectively mitigate forgetting and remarkably outperform the naíve combinations of previous art in video prediction and continual learning.

Transformers based on the attention mechanism have achieved impressive success in various areas. However, the attention mechanism has a quadratic complexity, significantly impeding Transformers from dealing with numerous tokens and scaling up to bigger models. Previous methods mainly utilize the similarity decomposition and the associativity of matrix multiplication to devise linear-time attention mechanisms. They avoid degeneration of attention to a trivial distribution by reintroducing inductive biases such as the locality, thereby at the expense of model generality and expressiveness. In this paper, we linearize Transformers free from specific inductive biases based on the flow network theory. We cast attention as the information flow aggregated from the sources (values) to the sinks (results) through the learned flow capacities (attentions). Within this framework, we apply the property of flow conservation into attention and propose the Flow-Attention mechanism of linear complexity. By respectively conserving the incoming flow of sinks for source competition and the outgoing flow of sources for sink allocation, Flow-Attention inherently generates informative attentions without using specific inductive biases. Empowered by the Flow-Attention, Flowformer yields strong performance in linear time for wide areas, including long sequence, time series, vision, natural language, and reinforcement learning. The code and settings are available at this repository: https://github.com/thuml/Flowformer.

Policy constraint methods to offline reinforcement learning (RL) typically utilize parameterization or regularization that constrains the policy to perform actions within the support set of the behavior policy. The elaborative designs of parameterization methods usually intrude into the policy networks, which may bring extra inference cost and cannot take full advantage of well-established online methods. Regularization methods reduce the divergence between the learned policy and the behavior policy, which may mismatch the inherent density-based definition of support set thereby failing to avoid the out-of-distribution actions effectively. This paper presents Supported Policy OpTimization (SPOT), which is directly derived from the theoretical formalization of the density-based support constraint. SPOT adopts a VAE-based density estimator to explicitly model the support set of behavior policy and presents a simple but effective density-based regularization term, which can be plugged non-intrusively into off-the-shelf off-policy RL algorithms. SPOT achieves the state-of-the-art performance on standard benchmarks for offline RL. Benefiting from the pluggable design, offline pretrained models from SPOT can also be applied to perform online fine-tuning seamlessly.

Deep neural networks achieve remarkable performances on a wide range of tasks with the aid of large-scale labeled datasets. Yet these datasets are time-consuming and labor-exhaustive to obtain on realistic tasks. To mitigate the requirement for labeled data, self-training is widely used in semi-supervised learning by iteratively assigning pseudo labels to unlabeled samples. Despite its popularity, self-training is well-believed to be unreliable and often leads to training instability. Our experimental studies further reveal that the bias in semi-supervised learning arises from both the problem itself and the inappropriate training with potentially incorrect pseudo labels, which accumulates the error in the iterative self-training process. To reduce the above bias, we propose Debiased Self-Training (DST). First, the generation and utilization of pseudo labels are decoupled by two parameter-independent classifier heads to avoid direct error accumulation. Second, we estimate the worst case of self-training bias, where the pseudo labeling function is accurate on labeled samples, yet makes as many mistakes as possible on unlabeled samples. We then adversarially optimize the representations to improve the quality of pseudo labels by avoiding the worst case. Extensive experiments justify that DST achieves an average improvement of 6.3% against state-of-the-art methods on standard semi-supervised learning benchmark datasets and 18.9% against FixMatch on 13 diverse tasks. Furthermore, DST can be seamlessly adapted to other self-training methods and help stabilize their training and balance performance across classes in both cases of training from scratch and finetuning from pre-trained models.

Transformers have shown great power in time series forecasting due to their global-range modeling ability. However, their performance can degenerate terribly on non-stationary real-world data in which the joint distribution changes over time. Previous studies primarily adopt stationarization to attenuate the non-stationarity of original series for better predictability. But the stationarized series deprived of inherent non-stationarity can be less instructive for real-world bursty events forecasting. This problem, termed over-stationarization in this paper, leads Transformers to generate indistinguishable temporal attentions for different series and impedes the predictive capability of deep models. To tackle the dilemma between series predictability and model capability, we propose Non-stationary Transformers as a generic framework with two interdependent modules: Series Stationarization and De-stationary Attention. Concretely, Series Stationarization unifies the statistics of each input and converts the output with restored statistics for better predictability. To address the over-stationarization problem, De-stationary Attention is devised to recover the intrinsic non-stationary information into temporal dependencies by approximating distinguishable attentions learned from raw series. Our Non-stationary Transformers framework consistently boosts mainstream Transformers by a large margin, which reduces MSE by 49.43% on Transformer, 47.34% on Informer, and 46.89% on Reformer, making them the state-of-the-art in time series forecasting. Code is available at this repository: https://github.com/thuml/Nonstationary_Transformers.

Transfer learning aims to leverage knowledge from pre-trained models to benefit the target task. Prior transfer learning work mainly transfers from a single model. However, with the emergence of deep models pre-trained from different resources, model hubs consisting of diverse models with various architectures, pre-trained datasets and learning paradigms are available. Directly applying single-model transfer learning methods to each model wastes the abundant knowledge of the model hub and suffers from high computational cost. In this paper, we propose a Hub-Pathway framework to enable knowledge transfer from a model hub. The framework generates data-dependent pathway weights, based on which we assign the pathway routes at the input level to decide which pre-trained models are activated and passed through, and then set the pathway aggregation at the output level to aggregate the knowledge from different models to make predictions. The proposed framework can be trained end-to-end with the target task-specific loss, where it learns to explore better pathway configurations and exploit the knowledge in pre-trained models for each target datum. We utilize a noisy pathway generator and design an exploration loss to further explore different pathways throughout the model hub. To fully exploit the knowledge in pre-trained models, each model is further trained by specific data that activate it, which ensures its performance and enhances knowledge transfer. Experiment results on computer vision and reinforcement learning tasks demonstrate that the proposed Hub-Pathway framework achieves the state-of-the-art performance for model hub transfer learning.

Model hubs with many pre-trained models (PTMs) have become a cornerstone of deep learning. Although built at a high cost, they remain under-exploited---practitioners usually pick one PTM from the provided model hub by popularity and then fine-tune the PTM to solve the target task. This naïve but common practice poses two obstacles to full exploitation of pre-trained model hubs: first, the PTM selection by popularity has no optimality guarantee, and second, only one PTM is used while the remaining PTMs are ignored. An alternative might be to consider all possible combinations of PTMs and extensively fine-tune each combination, but this would not only be prohibitive computationally but may also lead to statistical over-fitting. In this paper, we propose a new paradigm for exploiting model hubs that is intermediate between these extremes. The paradigm is characterized by two aspects: (1) We use an evidence maximization procedure to estimate the maximum value of label evidence given features extracted by pre-trained models. This procedure can rank all the PTMs in a model hub for various types of PTMs and tasks before fine-tuning. (2) The best ranked PTM can either be fine-tuned and deployed if we have no preference for the model's architecture or the target PTM can be tuned by the top $K$ ranked PTMs via a Bayesian procedure that we propose. This procedure, which we refer to as B-Tuning, not only improves upon specialized methods designed for tuning homogeneous PTMs, but also applies to the challenging problem of tuning heterogeneous PTMs where it yields a new level of benchmark performance.

This paper introduces video domain generalization where most video classification networks degenerate due to the lack of exposure to the target domains of divergent distributions. We observe that the global temporal features are less generalizable, due to the temporal domain shift that videos from other unseen domains may have an unexpected absence or misalignment of the temporal relations. This finding has motivated us to solve video domain generalization by effectively learning the local-relation features of different timescales that are more generalizable, and exploiting them along with the global-relation features to maintain the discriminability. This paper presents the VideoDG framework with two technical contributions. The first is a new deep architecture named the Adversarial Pyramid Network, which improves the generalizability of video features by capturing the local-relation, global-relation, and cross-relation features progressively. On the basis of pyramid features, the second contribution is a new and robust approach of adversarial data augmentation that can bridge different video domains by improving the diversity and quality of augmented data. We construct three video domain generalization benchmarks in which domains are divided according to different datasets, different consequences of actions, or different camera views, respectively. VideoDG consistently outperforms the combinations of previous video classification models and existing domain generalization methods on all benchmarks.

Unsupervised domain adaptation (UDA) enables transferring knowledge from a related source domain to a fully unlabeled target domain. Despite the significant advances in UDA, the performance gap remains quite large between UDA and supervised learning with fully labeled target data. Active domain adaptation (ADA) mitigates the gap under minimal annotation cost by selecting a small quota of target samples to annotate and incorporating them into training. Due to the domain shift, the query selection criteria of prior active learning methods may be ineffective to select the most informative target samples for annotation. In this paper, we propose Transferable Query Selection (TQS), which selects the most informative samples under domain shift by an ensemble of three new criteria: transferable committee, transferable uncertainty, and transferable domainness. We further develop a randomized selection algorithm to enhance the diversity of the selected samples. Experiments show that TQS remarkably outperforms previous UDA and ADA methods on several domain adaptation datasets. Deeper analyses demonstrate that TQS can select the most informative target samples under the domain shift.

Deep learning techniques for point clouds have achieved strong performance on a range of 3D vision tasks. However, it is costly to annotate large-scale point sets, making it critical to learn generalizable representations that can transfer well across different point sets. In this paper, we study a new problem of 3D Domain Generalization (3DDG) with the goal to generalize the model to other unseen domains of point clouds without any access to them in the training process. It is a challenging problem due to the substantial geometry shift from simulated to real data, such that most existing 3D models underperform due to overfitting the complete geometries in the source domain. We propose to tackle this problem via MetaSets, which meta-learns point cloud representations from a group of classification tasks on carefully-designed transformed point sets containing specific geometry priors. The learned representations are more generalizable to various unseen domains of different geometries. We design two benchmarks for Sim-to-Real transfer of 3D point clouds. Experimental results show that MetaSets outperforms existing 3D deep learning methods by large margins.

Domain adaptation (DA) aims at transferring knowledge from a labeled source domain to an unlabeled target domain. Though many DA theories and algorithms have been proposed, most of them are tailored into classification settings and may fail in regression tasks, especially in the practical keypoint detection task. To tackle this difficult but significant task, we present a method of regressive domain adaptation (RegDA) for unsupervised keypoint detection. Inspired by the latest theoretical work, we first utilize an adversarial regressor to maximize the disparity on the target domain and train a feature generator to minimize this disparity. However, due to the high dimension of the output space, this regressor fails to detect samples that deviate from the support of the source. To overcome this problem, we propose two important ideas. First, based on our observation that the probability density of the output space is sparse, we introduce a spatial probability distribution to describe this sparsity and then use it to guide the learning of the adversarial regressor. Second, to alleviate the optimization difficulty in the high-dimensional space, we innovatively convert the minimax game in the adversarial training to the minimization of two opposite goals. Extensive experiments show that our method brings large improvement by 8% to 11% in terms of PCK on different datasets.

Leveraging datasets available to learn a model with high generalization ability to unseen domains is important for computer vision, especially when the unseen domain’s annotated data are unavailable. We study a novel and practical problem of Open Domain Generalization (OpenDG), which learns from different source domains to achieve high performance on an unknown target domain, where the distributions and label sets of each individual source domain and the target domain can be different. The problem can be generally applied to diverse source domains and widely applicable to real-world applications. We propose a Domain-Augmented Meta-Learning framework to learn open-domain generalizable representations. We augment domains on both feature-level by a new Dirichlet mixup and label-level by distilled soft-labeling, which complements each domain with missing classes and other domain knowledge. We conduct meta-learning over domains by designing new meta-learning tasks and losses to preserve domain unique knowledge and generalize knowledge across domains simultaneously. Experiment results on various multi-domain datasets demonstrate that the proposed Domain-Augmented Meta-Learning (DAML) outperforms prior methods for unseen domain recognition.

This paper tackles video prediction from a new dimension of predicting spacetime-varying motions that are incessantly changing across both space and time. Prior methods mainly capture the temporal state transitions but overlook the complex spatiotemporal variations of the motion itself, making them difficult to adapt to ever-changing motions. We observe that physical world motions can be decomposed into transient variation and motion trend, while the latter can be regarded as the accumulation of previous motions. Thus, simultaneously capturing the transient variation and the motion trend is the key to make spacetime-varying motions more predictable. Based on these observations, we propose the MotionRNN framework, which can capture the complex variations within motions and adapt to spacetime-varying scenarios. MotionRNN has two main contributions. The first is that we design the MotionGRU unit, which can model the transient variation and motion trend in a unified way. The second is that we apply the MotionGRU to RNN-based predictive models and indicate a new flexible video prediction architecture with a Motion Highway, which can significantly improve the ability to predict changeable motions and avoid motion vanishing for stacked multiple-layer predictive models. With high flexibility, this framework can adapt to a series of models for deterministic spatiotemporal prediction. Our MotionRNN can yield significant improvements on three challenging benchmarks for video prediction with spacetime-varying motions.

Regression, as a counterpart to classification, is a major paradigm with a wide range of applications. Domain adaptation regression extends it by generalizing a regressor from a labeled source domain to an unlabeled target domain. Existing domain adaptation regression methods have achieved positive results limited only to the shallow regime. A question arises: Why learning invariant representations in the deep regime less pronounced? A key finding of this paper is that classification is robust to feature scaling but regression is not, and aligning the distributions of deep representations will alter feature scale and impede domain adaptation regression. Based on this finding, we propose to close the domain gap through orthogonal bases of the representation spaces, which are free from feature scaling. Inspired by Riemannian geometry of Grassmann manifold, we define a geometrical distance over representation subspaces and learn deep transferable representations by minimizing it. To avoid breaking the geometrical properties of deep representations, we further introduce the bases mismatch penalization to match the ordering of orthogonal bases across representation subspaces. Our method is evaluated on three domain adaptation regression benchmarks, two of which are introduced in this paper. Our method outperforms the state-of-the-art methods significantly, forming early positive results in the deep regime.

With the development of deep networks on various large-scale datasets, a large zoo of pretrained models are available. When transferring from a model zoo, applying classic single-model-based transfer learning methods to each source model suffers from high computational cost and cannot fully utilize the rich knowledge in the zoo. We propose \emph{Zoo-Tuning} to address these challenges, which learns to adaptively transfer the parameters of pretrained models to the target task. With the learnable channel alignment layer and adaptive aggregation layer, Zoo-Tuning \emph{adaptively aggregates channel aligned pretrained parameters to derive the target model}, which simultaneously promotes knowledge transfer and adapts source models to downstream tasks. The adaptive aggregation substantially reduces the computation cost at both training and inference. We further propose lite Zoo-Tuning with the temporal ensemble of batch average gating values to reduce the storage cost at the inference time. We evaluate our approach on a variety of tasks, including reinforcement learning, image classification, and facial landmark detection. Experiment results demonstrate that the proposed adaptive transfer learning approach can more effectively and efficiently transfer knowledge from a zoo of models.

Deep learning has made revolutionary advances to diverse applications in the presence of large-scale labeled datasets. However, it is prohibitively time-costly and labor-expensive to collect sufficient labeled data in most realistic scenarios. To mitigate the requirement for labeled data, semi-supervised learning (SSL) focuses on simultaneously exploring both labeled and unlabeled data, while transfer learning (TL) popularizes a favorable practice of fine-tuning a pre-trained model to the target data. A dilemma is thus encountered: Without a decent pre-trained model to provide an implicit regularization, SSL through self-training from scratch will be easily misled by inaccurate pseudo-labels, especially in large-sized label space; Without exploring the intrinsic structure of unlabeled data, TL through fine-tuning from limited labeled data is at risk of under-transfer caused by model shift. To escape from this dilemma, we present Self-Tuning to enable data-efficient deep learning by unifying the exploration of labeled and unlabeled data and the transfer of a pre-trained model, as well as a Pseudo Group Contrast (PGC) mechanism to mitigate the reliance on pseudo-labels and boost the tolerance to false labels. Self-Tuning outperforms its SSL and TL counterparts on five tasks by sharp margins, e.g. it doubles the accuracy of fine-tuning on Cars with $15%$ labels.

This paper studies task adaptive pre-trained model selection, an underexplored problem of assessing pre-trained models for the target task and select best ones from the model zoo \emph{without fine-tuning}. A few pilot works addressed the problem in transferring supervised pre-trained models to classification tasks, but they cannot handle emerging unsupervised pre-trained models or regression tasks. In pursuit of a practical assessment method, we propose to estimate the maximum value of label evidence given features extracted by pre-trained models. Unlike the maximum likelihood, the maximum evidence is \emph{immune to over-fitting}, while its expensive computation can be dramatically reduced by our carefully designed algorithm. The Logarithm of Maximum Evidence (LogME) can be used to assess pre-trained models for transfer learning: a pre-trained model with a high LogME value is likely to have good transfer performance. LogME is \emph{fast, accurate, and general}, characterizing itself as the first practical method for assessing pre-trained models. Compared with brute-force fine-tuning, LogME brings at most $3000\times$ speedup in wall-clock time and requires only $1%$ memory footprint. It outperforms prior methods by a large margin in their setting and is applicable to new settings. It is general enough for diverse pre-trained models (supervised pre-trained and unsupervised pre-trained), downstream tasks (classification and regression), and modalities (vision and language). Code is available at this repository: \href{https://github.com/thuml/LogME}{https://github.com/thuml/LogME}.

Mainstream approaches for unsupervised domain adaptation (UDA) learn domain-invariant representations to narrow the domain shift, which are empirically effective but theoretically challenged by the hardness or impossibility theorems. Recently, self-training has been gaining momentum in UDA, which exploits unlabeled target data by training with target pseudo-labels. However, as corroborated in this work, under distributional shift, the pseudo-labels can be unreliable in terms of their large discrepancy from target ground truth. In this paper, we propose Cycle Self-Training (CST), a principled self-training algorithm that explicitly enforces pseudo-labels to generalize across domains. CST cycles between a forward step and a reverse step until convergence. In the forward step, CST generates target pseudo-labels with a source-trained classifier. In the reverse step, CST trains a target classifier using target pseudo-labels, and then updates the shared representations to make the target classifier perform well on the source data. We introduce the Tsallis entropy as a confidence-friendly regularization to improve the quality of target pseudo-labels. We analyze CST theoretically under realistic assumptions, and provide hard cases where CST recovers target ground truth, while both invariant feature learning and vanilla self-training fail. Empirical results indicate that CST significantly improves over the state-of-the-arts on visual recognition and sentiment analysis benchmarks.

Extending the forecasting time is a critical demand for real applications, such as extreme weather early warning and long-term energy consumption planning. This paper studies the long-term forecasting problem of time series. Prior Transformer-based models adopt various self-attention mechanisms to discover the long-range dependencies. However, intricate temporal patterns of the long-term future prohibit the model from finding reliable dependencies. Also, Transformers have to adopt the sparse versions of point-wise self-attentions for long series efficiency, resulting in the information utilization bottleneck. Going beyond Transformers, we design Autoformer as a novel decomposition architecture with an Auto-Correlation mechanism. We break with the pre-processing convention of series decomposition and renovate it as a basic inner block of deep models. This design empowers Autoformer with progressive decomposition capacities for complex time series. Further, inspired by the stochastic process theory, we design the Auto-Correlation mechanism based on the series periodicity, which conducts the dependencies discovery and representation aggregation at the sub-series level. Auto-Correlation outperforms self-attention in both efficiency and accuracy. In long-term forecasting, Autoformer yields state-of-the-art accuracy, with a 38% relative improvement on six benchmarks, covering five practical applications: energy, traffic, economics, weather and disease. Code is available at this repository: https://github.com/thuml/Autoformer.

Cross-domain sentiment classification aims to leverage useful knowledge from a source domain to mitigate the supervision sparsity in a target domain. A series of approaches depend on the pivot features that behave similarly for polarity prediction in both domains. However, the engineering of such pivot features remains cumbersome and prevents us from learning the disentangled and transferable representations from rich semantic and syntactic information. Towards learning the pivots and representations simultaneously, we propose a new Transferable Pivot Transformer (TPT). Our model consists of two networks: a Pivot Selector that learns to detect transferable n-gram pivots from contexts, and a Transferable Transformer that learns to generate domain-invariant representations by modeling the correlation between pivot and non-pivot words. The Pivot Selector and Transferable Transformer are jointly optimized through end-to-end back-propagation. We experiment with real tasks of cross-domain sentiment classification over 20 domain pairs where our model outperforms prior arts.

We study a new research problem of probabilistic future frames prediction from a sequence of noisy inputs, which is useful because it is difficult to guarantee the quality of input frames in practical spatiotemporal prediction applications. It is also challenging because it involves two levels of uncertainty: the perceptual uncertainty from noisy observations and the dynamics uncertainty in forward modeling. In this paper, we propose to tackle this problem with an end-to-end trainable model named Bayesian Predictive Network (BP-Net). Unlike previous work in stochastic video prediction that assumes spatiotemporal coherence and therefore fails to deal with perceptual uncertainty, BP-Net models both levels of uncertainty in an integrated framework. Furthermore, unlike previous work that can only provide unsorted estimations of future frames, BP-Net leverages a differentiable sequential importance sampling (SIS) approach to make future predictions based on the inference of underlying physical states, thereby providing sorted prediction candidates in accordance with the SIS importance weights, i.e., the confidences. Our experiment results demonstrate that BP-Net remarkably outperforms existing approaches on predicting future frames from noisy data.

Fine-grained visual categorization has long been considered as an important problem, however, its real application is still restricted, since precisely annotating a large fine-grained image dataset is a laborious task and requires expert-level human knowledge. A solution to this problem is applying domain adaptation approaches to fine-grained scenarios, where the key idea is to discover the commonality between existing fine-grained image datasets and massive unlabeled data in the wild. The main technical bottleneck lies in that the large inter-domain variation will deteriorate the subtle boundaries of small inter-class variation during domain alignment. This paper presents the Progressive Adversarial Networks (PAN) to align fine-grained categories across domains with a curriculum-based adversarial learning framework. In particular, throughout the learning process, domain adaptation is carried out through all multi-grained features, progressively exploiting the label hierarchy from coarse to fine. The progressive learning is applied upon both category classification and domain alignment, boosting both the discriminability and the transferability of the fine-grained features. Our method is evaluated on three benchmarks, two of which are proposed by us, and it outperforms the state-of-the-art domain adaptation methods.

While multi-discriminators have been recently exploited to enhance the discriminability and diversity of Generative Adversarial Networks (GANs), these independent discriminators may not collaborate harmoniously to learn diverse and complementary decision boundaries. This paper extends the original two-player adversarial game of GANs by introducing a new multi-player objective named Discriminator Discrepancy Loss (DDL) for diversifying the multi-discriminators. Besides the competition between the generator and each discriminator, there are also competitions between the discriminators: 1) When training multi-discriminators, we simultaneously minimize the original GAN loss and maximize DDL, seeking a good trade-off between the accuracy and diversity. This yields diversified multi-discriminators that fit the generated data distribution to the real data distribution from more comprehensive perspectives. 2) When training the generator, we minimize DDL to encourage the generator to confuse all discriminators. This enhances the diversity of the generated data distribution. Further, we propose a layer-sharing network architecture for the multi-discriminators, which allows them to learn from distinct perspectives about the shared low-level features through better collaboration. It also makes our model more lightweight than existing multi-discriminators approaches. Our DDL-GAN remarkably outperforms other GANs over five standard datasets for image generation tasks.

In classic video action recognition, labels may not contain enough information about the diverse video appearance and dynamics, thus, existing models that are trained under the standard supervised learning paradigm may extract less generalizable features. We evaluate these models under a cross-dataset experiment setting, as the above label bias problem in video analysis is even more prominent across different data sources. We find that using the optical flows as model inputs harms the generalization ability of most video recognition models.Based on these findings, we present a multi-task learning paradigm for video classification. Our key idea is to avoid label bias and improve the generalization ability by taking data as its own supervision or supervising constraints on the data. First, we take the optical flows and the RGB frames by taking them as auxiliary supervisions, and thus naming our model as Reversed Two-Stream Networks (Rev2Net). Further, we collaborate the auxiliary flow prediction task and the frame reconstruction task by introducing a new training objective to Rev2Net, named Decoding Discrepancy Penalty (DDP), which constraints the discrepancy of the multi-task features in a self-supervised manner. Rev2Net is shown to be effective on the classic action recognition task. It specifically shows a strong generalization ability in the cross-dataset experiments.

This paper explores a new research problem of unsupervised transfer learning across multiple spatiotemporal prediction tasks. Unlike most existing transfer learning methods that focus on fixing the discrepancy between supervised tasks, we study how to transfer knowledge from a zoo of unsupervisedly learned models towards another predictive network. Our motivation is that models from different sources are expected to understand the complex spatiotemporal dynamics from different perspectives, thereby effectively supplementing the new task, even if the task has sufficient training samples. Technically, we propose a differentiable framework named transferable memory. It adaptively distills knowledge from a bank of memory states of multiple pretrained RNNs, and applies it to the target network via a novel recurrent structure called the Transferable Memory Unit (TMU). Compared with finetuning, our approach yields significant improvements on three benchmarks for spatiotemporal prediction, and benefits the target task even from less relevant pretext ones.

Domain Adaptation (DA) aims at transferring knowledge from a labeled source domain to an unlabeled target domain. While re- markable advances have been witnessed recently, the power of DA methods still heavily depends on the network depth, especially when the domain discrepancy is large, posing an unprecedented challenge to DA in low-resource scenarios where fast and adaptive inference is required. How to bridge transferability and resource- efficient inference in DA becomes an important problem. In this paper, we propose Resource Efficient Domain Adaptation (REDA), a general framework that can adaptively adjust computation re- sources across 'easier' and 'harder' inputs. Based on existing multi- exit architectures, REDA has two novel designs: 1) Transferable distillation to distill the transferability of top classifier into the early exits; 2) Consistency weighting to control the distillation degree via prediction consistency. As a general method, REDA can be easily applied with a variety of DA methods. Empirical results and analy- ses justify that REDA can substantially improve the accuracy and accelerate the inference under domain shift and low resource.

Fine-tuning pre-trained deep networks on a small dataset is an important component in the deep learning pipeline. A critical problem in fine-tuning is how to avoid over-fitting when data are limited. Existing efforts work from two aspects: (1) impose regularization on parameters or features; (2) transfer prior knowledge to fine-tuning by reusing pre-trained parameters. In this paper, we take an alternative approach by refactoring the widely used Batch Normalization (BN) module to mitigate over-fitting. We propose a two-branch design with one branch normalized by mini-batch statistics and the other branch normalized by moving statistics. During training, two branches are stochastically selected to avoid over-depending on some sample statistics, resulting in a strong regularization effect, which we interpret as ``architecture regularization.'' The resulting method is dubbed stochastic normalization (\textbf{StochNorm}). With the two-branch architecture, it naturally incorporates pre-trained moving statistics in BN layers during fine-tuning, exploiting more prior knowledge of pre-trained networks. Extensive empirical experiments show that StochNorm is a powerful tool to avoid over-fitting in fine-tuning with small datasets. Besides, StochNorm is readily pluggable in modern CNN backbones. It is complementary to other fine-tuning methods and can work together to achieve stronger regularization effect.

Domain adaptation aims at knowledge transfer from a labeled source domain to an unlabeled target domain. Current domain adaptation methods have made substantial advances in adapting discrete domains. However, this can be unrealistic in real-world applications, where target data usually comes in an online and continually evolving manner as small batches, posing challenges to classic domain adaptation paradigm: (1) Mainstream domain adaptation methods are tailored to stationary target domains, and can fail in non-stationary environments. (2) Since the target data arrive online, the agent should also maintain competence on previous target domains, i.e. to adapt without forgetting. To tackle these challenges, we propose a meta-adaptation framework which enables the learner to adapt to continually evolving target domain without catastrophic forgetting. Our framework comprises of two components: a meta-objective of learning representations to adapt to evolving domains, enabling meta-learning for unsupervised domain adaptation; and a meta-adapter for learning to adapt without forgetting, reserving knowledge from previous target data. Experiments validate the effectiveness our method on evolving target domains.

Domain Adaptation (DA) enables transferring a learning machine from a labeled source domain to an unlabeled target one. While remarkable advances have been made, most of the existing DA methods focus on improving the target accuracy at inference. How to estimate the predictive uncertainty of DA models is vital for decision-making in safety-critical scenarios but remains the boundary to explore. In this paper, we delve into the open problem of Calibration in DA, which is extremely challenging due to the coexistence of domain shift and the lack of target labels. We first reveal the dilemma that DA models learn higher accuracy at the expense of well-calibrated probabilities. Driven by this finding, we propose Transferable Calibration (TransCal) to achieve more accurate calibration with lower bias and variance in a unified hyperparameter-free optimization framework. As a general post-hoc calibration method, TransCal can be easily applied to recalibrate existing DA methods. Its efficacy has been justified both theoretically and empirically.

Fine-tuning pre-trained deep neural networks (DNNs) to a target dataset, also known as transfer learning, is widely used in computer vision and NLP. Because task-specific layers mainly contain categorical information and categories vary with datasets, practitioners only \textit{partially} transfer pre-trained models by discarding task-specific layers and fine-tuning bottom layers. However, it is a reckless loss to simply discard task-specific parameters who take up as many as $20\%$ of the total parameters in pre-trained models. To \textit{fully} transfer pre-trained models, we propose a two-step framework named \textbf{Co-Tuning}: (i) learn the relationship between source categories and target categories from the pre-trained model and calibrated predictions; (ii) target labels (one-hot labels), as well as source labels (probabilistic labels) translated by the category relationship, collaboratively supervise the fine-tuning process. A simple instantiation of the framework shows strong empirical results in four visual classification tasks and one NLP classification task, bringing up to $20\%$ relative improvement. While state-of-the-art fine-tuning techniques mainly focus on how to impose regularization when data are not abundant, Co-Tuning works not only in medium-scale datasets (100 samples per class) but also in large-scale datasets (1000 samples per class) where regularization-based methods bring no gains over the vanilla fine-tuning. Co-Tuning relies on a typically valid assumption that the pre-trained dataset is diverse enough, implying its broad application area.

Domain adaptation improves a target task by knowledge transfer from a source domain with rich annotations. It is not uncommon that “source-domain engineering” becomes a cumbersome process in domain adaptation: the high-quality source domains highly related to the target domain are hardly available. Thus, weakly-supervised domain adaptation has been introduced to address this difficulty, where we can tolerate the source domain with noises in labels, features, or both. As such, for a particular target task, we simply collect the source domain with coarse labeling or corrupted data. In this paper, we try to address two entangled challenges of weaklysupervised domain adaptation: sample noises of the source domain and distribution shift across domains. To disentangle these challenges, a Transferable Curriculum Learning (TCL) approach is proposed to train the deep networks, guided by a transferable curriculum informing which of the source examples are noiseless and transferable. The approach enhances positive transfer from clean source examples to the target and mitigates negative transfer of noisy source examples. A thorough evaluation shows that our approach significantly outperforms the state-of-the-art on weakly-supervised domain adaptation tasks.

Recent work in domain adaptation bridges different domains by adversarially learning a domain-invariant representation that cannot be distinguished by a domain discriminator. Existing methods of adversarial domain adaptation mainly align the global images across the source and target domains. However, it is obvious that not all regions of an image are transferable, while forcefully aligning the untransferable regions may lead to negative transfer. Furthermore, some of the images are significantly dissimilar across domains, resulting in weak image-level transferability. To this end, we present Transferable Attention for Domain Adaptation (TADA), focusing our adaptation model on transferable regions or images. We implement two types of complementary transferable attention: transferable local attention generated by multiple region-level domain discriminators to highlight transferable regions, and transferable global attention generated by single image-level domain discriminator to highlight transferable images. Extensive experiments validate that our proposed models exceed state of the art results on standard domain adaptation datasets.

Domain adaptation is critical for learning in new and unseen environments. With domain adversarial training, deep networks can learn disentangled and transferable features that effectively diminish the dataset shift between the source and target domains for knowledge transfer. In the era of Big Data, large-scale labeled datasets are readily available, stimulating the interest in partial domain adaptation (PDA), which transfers a recognizer from a large labeled domain to a small unlabeled domain. It extends standard domain adaptation to the scenario where target labels are only a subset of source labels. Under the condition that target labels are unknown, the key challenges of PDA are how to transfer relevant examples in the shared classes to promote positive transfer and how to ignore irrelevant ones in the source domain to mitigate negative transfer. In this work, we propose a unified approach to PDA, Example Transfer Network (ETN), which jointly learns domain-invariant representations across domains and a progressive weighting scheme to quantify the transferability of source examples. A thorough evaluation on several benchmark datasets shows that ETN consistently achieves state-of-the-art results for various partial domain adaptation tasks.

Domain adaptation has become a resounding success in leveraging labeled data from a source domain to learn an accurate classifier for an unlabeled target domain. When deployed in the wild, the target domain usually contains unknown classes that are not observed in the source domain. Such setting is termed Open Set Domain Adaptation (OSDA). While several methods have been proposed to address OSDA, none of them takes into account the openness of the target domain, which is measured by the proportion of unknown classes in all target classes. Openness is a critical point in open set domain adaptation and exerts a significant impact on performance. In addition, current work aligns the entire target domain with the source domain without excluding unknown samples, which may give rise to negative transfer due to the mismatch between unknown and known classes. To this end, this paper presents Separate to Adapt (STA), an end-to-end approach to open set domain adaptation. The approach adopts a coarse-to-fine weighting mechanism to progressively separate the samples of unknown and known classes, and simultaneously weigh their importance on feature distribution alignment. Our approach allows openness-robust open set domain adaptation, which can be adaptive to a variety of openness in the target domain. We evaluate STA on several benchmark datasets of various openness levels. Results verify that STA significantly outperforms previous methods.

Natural spatiotemporal processes can be highly non-stationary in many ways, e.g. the low-level non-stationarity such as spatial correlations or temporal dependencies of local pixel values; and the high-level variations such as the accumulation, deformation or dissipation of radar echoes in precipitation forecasting. From Cramer's Decomposition, any non-stationary process can be decomposed into deterministic, time-variant polynomials, plus a zero-mean stochastic term. By applying differencing operations appropriately, we may turn time-variant polynomials into a constant, making the deterministic component predictable. However, most previous recurrent neural networks for spatiotemporal prediction do not use the differential signals effectively, and their relatively simple state transition functions prevent them from learning too complicated variations in spacetime. We propose the Memory In Memory (MIM) networks and corresponding recurrent blocks for this purpose. The MIM blocks exploit the differential signals between adjacent recurrent states to model the non-stationary and approximately stationary properties in spatiotemporal dynamics with two cascaded, self-renewed memory modules. By stacking multiple MIM blocks, we could potentially handle higher-order non-stationarity. The MIM networks achieve the state-of-the-art results on four spatiotemporal prediction tasks across both synthetic and real-world datasets. We believe that the general idea of this work can be potentially applied to other time-series forecasting tasks.

Domain adaptation aims to transfer knowledge in the presence of the domain gap. Existing domain adaptation methods rely on rich prior knowledge about the relationship between the label sets of source and target domains, which greatly limits their application in the wild. This paper introduces Universal Domain Adaptation (UDA) that requires no prior knowledge on the label sets. For a given source label set and a target label set, they may contain a common label set and hold a private label set respectively, bringing up an additional category gap. UDA requires a model to either (1) classify the target sample correctly if it is associated with a label in the common label set, or (2) mark it as ``unknown'' otherwise. More importantly, a UDA model should work stably against a wide spectrum of commonness (the proportion of the common label set over the complete label set) so that it can handle real-world problems with unknown target label sets. To solve the universal domain adaptation problem, we propose Universal Adaptation Network (UAN). It quantifies sample-level transferability to discover the common label set and the label sets private to each domain, thereby promoting the adaptation in the automatically discovered common label set and recognizing the ``unknown'' samples successfully. A thorough evaluation shows that UAN outperforms the state of the art closed set, partial and open set domain adaptation methods in the novel UDA setting.

Deep hashing enables computation and memory efficient image search through end-to-end learning of feature representations and binary codes. While linear scan over binary hash codes is more efficient than over the high-dimensional representations, its linear-time complexity is still unacceptable for very large databases. Hamming space retrieval enables constant-time search through hash lookups, where for each query, there is a Hamming ball centered at the query and the data points within the ball are returned as relevant. Since inside the Hamming ball implies retrievable while outside irretrievable, it is crucial to explicitly characterize the Hamming ball. The main idea of this work is to directly embody the Hamming radius into the loss functions, leading to Maximum-Margin Hamming Hashing (MMHH), a new model specifically optimized for Hamming space retrieval. We introduce a max-margin t-distribution loss, where the t-distribution concentrates more similar data points to be within the Hamming ball, and the margin characterizes the Hamming radius such that less penalization is applied to similar data points within the Hamming ball. The loss function also introduces robustness to data noise, where the similarity supervision may be inaccurate in practical problems. The model is trained end-to-end using a new semi-batch optimization algorithm tailored to extremely imbalanced data. Our method yields state-of-the-art results on four datasets and shows superior performance on noisy data.

We present a Z-Order RNN (Znet) for predicting future video frames given historical observations. There are two main contributions respectively in deterministic and stochastic modeling perspective. First, we propose a new RNN architecture for modeling the deterministic dynamics, which updates hidden states along a z-order curve to enhance the consistency of the features of mirrored layers. Second, we introduce an adversarial training approach to two-stream Znet for modeling the stochastic variations, which forces the Znet-Predictor to imitate the behavior of the Znet-Probe. This two-stream architecture enables the adversarial training to be conducted in the feature space instead of the image space. Our model achieves the state-of-the-art prediction accuracy on two video datasets.

This paper addresses the problem of unsupervised domain adaption from theoretical and algorithmic perspectives. Existing domain adaptation theories naturally imply minimax optimization algorithms, which connect well with the domain adaptation methods based on adversarial learning. However, several disconnections still exist and form the gap between theory and algorithm. We extend previous theories (Mansour et al., 2009c; Ben-David et al., 2010) to multiclass classification in domain adaptation, where classifiers based on the scoring functions and margin loss are standard choices in algorithm design. We introduce Margin Disparity Discrepancy, a novel measurement with rigorous generalization bounds, tailored to the distribution comparison with the asymmetric margin loss, and to the minimax optimization for easier training. Our theory can be seamlessly transformed into an adversarial learning algorithm for domain adaptation, successfully bridging the gap between theory and algorithm. A series of empirical studies show that our algorithm achieves the state of the art accuracies on challenging domain adaptation tasks.

Adversarial domain adaptation has made remarkable advances in learning transferable representations for knowledge transfer across domains. While adversarial learning strengthens the feature transferability which the community focuses on, its impact on the feature discriminability has not been fully explored. In this paper, a series of experiments based on spectral analysis of the feature representations have been conducted, revealing an unexpected deterioration of the discriminability while learning transferable features adversarially. Our key finding is that the eigenvectors with the largest singular values will dominate the feature transferability. As a consequence, the transferability is enhanced at the expense of over penalization of other eigenvectors that embody rich structures crucial for discriminability. Towards this problem, we present Batch Spectral Penalization (BSP), a general approach to penalizing the largest singular values so that other eigenvectors can be relatively strengthened to boost the feature discriminability. Experiments show that the approach significantly improves upon representative adversarial domain adaptation methods to yield state of the art results.

Domain adaptation enables knowledge transfer from a labeled source domain to an unlabeled target domain. A mainstream approach is adversarial feature adaptation, which learns domain-invariant representations through aligning the feature distributions of both domains. However, a theoretical prerequisite of domain adaptation is the adaptability measured by the expected risk of an ideal joint hypothesis over the source and target domains. In this respect, adversarial feature adaptation may potentially deteriorate the adaptability, since it distorts the original feature distributions when suppressing domain-specific variations. To this end, we propose Transferable Adversarial Training (TAT) to enable the adaptation of deep classifiers. The approach generates transferable examples to fill in the gap between the source and target domains, and adversarially trains the deep classifiers to make consistent predictions over the transferable examples. Without learning domain-invariant representations at the expense of distorting the feature distributions, the adaptability in the theoretical learning bound is algorithmically guaranteed. A series of experiments validate that our approach advances the state of the arts on a variety of domain adaptation tasks in vision and NLP, including object recognition, learning from synthetic to real data, and sentiment classification.

Deep unsupervised domain adaptation (Deep UDA) methods successfully leverage rich labeled data in a source domain to boost the performance on related but unlabeled data in a target domain. However, algorithm comparison is cumbersome in Deep UDA due to the absence of accurate and standardized model selection method, posing an obstacle to further advances in the field. Existing model selection methods for Deep UDA are either highly biased, restricted, unstable, or even controversial (requiring labeled target data). To this end, we propose Deep Embedded Validation (DEV), which embeds adapted feature representation into the validation procedure to obtain unbiased estimation of the target risk with bounded variance. The variance is further reduced by the technique of control variate. The efficacy of the method has been justified both theoretically and empirically.

Before sufficient training data is available, fine-tuning neural networks pre-trained on large-scale datasets substantially outperforms training from random initialization. However, fine-tuning methods suffer from two dilemmas, catastrophic forgetting and negative transfer. While several methods with explicit attempts to overcome catastrophic forgetting have been proposed, negative transfer is rarely delved into. In this paper, we launch an in-depth empirical investigation into negative transfer in fine-tuning and find that, for the weight parameters and feature representations, transferability of their spectral components is diverse. For safe transfer learning, we present Batch Spectral Shrinkage (BSS), a novel regularization approach to penalizing smaller singular values so that untransferable spectral components are suppressed. BSS is orthogonal to existing fine-tuning methods and is readily pluggable to them. Experimental results show that BSS can significantly enhance the performance of representative methods, especially with limited training data.

Deep neural networks (DNNs) excel at learning representations when trained on large-scale datasets. Pre-trained DNNs also show strong transferability when fine-tuned to other labeled datasets. However, such transferability becomes weak when the target dataset is fully unlabeled as in Unsupervised Domain Adaptation (UDA). We envision that the loss of transferability may stem from the intrinsic limitation of the architecture design of DNNs. In this paper, we delve into the components of DNN architectures and propose Transferable Normalization (TransNorm) in place of existing normalization techniques. TransNorm is an end-to-end trainable layer to make DNNs more transferable across domains. As a general method, TransNorm can be easily applied to various deep neural networks and domain adaption methods, without introducing any extra hyper-parameters or learnable parameters. Empirical results justify that TransNorm not only improves classification accuracies but also accelerates convergence for mainstream DNN-based domain adaptation methods.

Domain adaptation studies learning algorithms that generalize across source domains and target domains that exhibit different distributions. Recent studies reveal that deep neural networks can learn transferable features that generalize well to similar novel tasks. However, as deep features eventually transition from general to specific along the network, feature transferability drops significantly in higher task-specific layers with increasing domain discrepancy. To formally reduce the effects of this discrepancy and enhance feature transferability in task-specific layers, we develop a novel framework for deep adaptation networks that extends deep convolutional neural networks to domain adaptation problems. The framework embeds the deep features of all task-specific layers into reproducing kernel Hilbert spaces (RKHSs) and optimally matches different domain distributions. The deep features are made more transferable by exploiting low-density separation of target-unlabeled data in very deep architectures, while the domain discrepancy is further reduced via the use of multiple kernel learning that enhances the statistical power of kernel embedding matching. The overall framework is cast in a minimax game setting. Extensive empirical evidence shows that the proposed networks yield state-of-the-art results on standard visual domain-adaptation benchmarks.

Hashing is widely applied to large-scale image retrieval due to the storage and retrieval efficiency. Existing work on deep hashing assumes that the database in the target domain is identically distributed with the training set in the source domain. This paper relaxes this assumption to a transfer retrieval setting, which allows the database and the training set to come from different but relevant domains. However, the transfer retrieval setting will introduce two technical difficulties: first, the hash model trained on the source domain cannot work well on the target domain due to the large distribution gap; second, the domain gap makes it difficult to concentrate the database points to be within a small Hamming ball. As a consequence, transfer retrieval performance within Hamming Radius 2 degrades significantly in existing hashing methods. This paper presents Transfer Adversarial Hashing (TAH), a new hybrid deep architecture that incorporates a pairwise t-distribution cross-entropy loss to learn concentrated hash codes and an adversarial network to align the data distributions between the source and target domains. TAH can generate compact transfer hash codes for efficient image retrieval on both source and target domains. Comprehensive experiments validate that TAH yields state of the art Hamming space retrieval performance on standard datasets.

Domain adaptation generalizes a learning model across source domain and target domain that follow different distributions. Most existing work follows a two-step procedure: first, explores either feature matching or instance reweighting independently, and second, train the transfer classifier separately. In this paper, we show that either feature matching or instance reweighting can only reduce, but not remove, the cross-domain discrepancy, and the knowledge hidden in the relations between the data labels from the source and target domains is important for unsupervised domain adaptation. We propose a new Distribution Matching Machine (DMM) based on the structural risk minimization principle, which learns a transfer support vector machine by extracting invariant feature representations and estimating unbiased instance weights that jointly minimize the cross-domain distribution discrepancy. This leads to a robust transfer learner that performs well against both mismatched features and irrelevant instances. Our theoretical analysis proves that the proposed approach further reduces the generalization error bound of related domain adaptation methods. Comprehensive experiments validate that the DMM approach significantly outperforms competitive methods on standard domain adaptation benchmarks.

Recent advances in deep domain adaptation reveal that adversarial learning can be embedded into deep networks to learn transferable features that reduce distribution discrepancy between the source and target domains. Existing domain adversarial adaptation methods based on single domain discriminator only align the source and target data distributions without exploiting the complex multimode structures. In this paper, we present a multi-adversarial domain adaptation (MADA) approach, which captures multimode structures to enable fine-grained alignment of different data distributions based on multiple domain discriminators. The adaptation can be achieved by stochastic gradient descent with the gradients computed by back-propagation in linear-time. Empirical evidence demonstrates that the proposed model outperforms state of the art methods on standard domain adaptation datasets.

Adversarial learning has been successfully embedded into deep networks to learn transferable features, which reduce distribution discrepancy between the source and target domains. Existing domain adversarial networks assume fully shared label space across domains. In the presence of big data, there is strong motivation of transferring both classification and representation models from existing large-scale domains to unknown small-scale domains. This paper introduces partial transfer learning, which relaxes the shared label space assumption to that the target label space is only a subspace of the source label space. Previous methods typically match the whole source domain to the target domain, which are prone to negative transfer for the partial transfer problem. We present Selective Adversarial Network (SAN), which simultaneously circumvents negative transfer by selecting out the outlier source classes and promotes positive transfer by maximally matching the data distributions in the shared label space. Experiments demonstrate that our models exceed state-of-the-art results for partial transfer learning tasks on several benchmark datasets.

Due to its computation efficiency and retrieval quality, hashing has been widely applied to approximate nearest neighbor search for large-scale image retrieval, while deep hashing further improves the retrieval quality by end-to-end representation learning and hash coding. With compact hash codes, Hamming space retrieval enables the most efficient constant-time search that returns data points within a given Hamming radius to each query, by hash table lookups instead of linear scan. However, subject to the weak capability of concentrating relevant images to be within a small Hamming ball due to mis-specified loss functions, existing deep hashing methods may underperform for Hamming space retrieval. This work presents Deep Cauchy Hashing (DCH), a novel deep hashing model that generates compact and concentrated binary hash codes to enable efficient and effective Hamming space retrieval. The main idea is to design a pairwise cross-entropy loss based on Cauchy distribution, which penalizes significantly on similar image pairs with Hamming distance larger than the given Hamming radius threshold. Comprehensive experiments demonstrate that DCH can generate highly concentrated hash codes and yield state-of-the-art Hamming space retrieval performance on three datasets, NUS-WIDE, CIFAR-10, and MS-COCO.

Deep learning to hash improves image retrieval performance by end-to-end representation learning and hash coding from training data with pairwise similarity information. Subject to the scarcity of similarity information that is often expensive to collect for many application domains, existing deep learning to hash methods may overfit the training data and result in substantial loss of retrieval quality. This paper presents HashGAN, a novel architecture for deep learning to hash, which learns compact binary hash codes from both real images and diverse images synthesized by generative models. The main idea is to augment the training data with nearly real images synthesized from a new Pair Conditional Wasserstein GAN (PC-WGAN) conditioned on the pairwise similarity information. Extensive experiments demonstrate that HashGAN can generate high-quality binary hash codes and yield state-of-the-art image retrieval performance on three benchmarks, NUS-WIDE, CIFAR-10, and MS-COCO.

We present PredRNN++, a recurrent network for spatiotemporal predictive learning. In pursuit of a great modeling capability for short-term video dynamics, we make our network deeper in time by leveraging a new recurrent structure named Causal LSTM with cascaded dual memories. To alleviate the gradient propagation difficulties in deep predictive models, we propose a Gradient Highway Unit, which provides alternative quick routes for the gradient flows from outputs back to long-range previous inputs. The gradient highway units work seamlessly with the causal LSTMs, enabling our model to capture the short-term and the long-term video dependencies adaptively. Our model achieves state-of-the-art prediction results on both synthetic and real video datasets, showing its power in modeling entangled motions.

Predicting future frames in videos remains an unsolved but challenging problem. Mainstream recurrent models suffer from huge memory usage and computation cost, while convolutional models are unable to effectively capture the temporal dependencies between consecutive video frames. To tackle this problem, we introduce an entirely CNN-based architecture, PredCNN, that models the dependencies between the next frame and the sequential video inputs. Inspired by the core idea of recurrent models that previous states have more transition operations than future states, we design a cascade multiplicative unit (CMU) that provides relatively more operations for previous video frames. This newly proposed unit enables PredCNN to predict future spatiotemporal data without any recurrent chain structures, which eases gradient propagation and enables a fully paralleled optimization. We show that PredCNN outperforms the state-of-the-art recurrent models for video prediction on the standard Moving MNIST dataset and two challenging crowd flow prediction datasets, and achieves a faster training speed and lower memory footprint.

Deep hashing enables image retrieval by end-to-end learning of deep representations and hash codes from training data with pairwise similarity information. Subject to the distribution skewness underlying the similarity information, most existing deep hashing methods may underperform for imbalanced data due to misspecified loss functions. This paper presents Deep Priority Hashing (DPH), an end-to-end architecture that generates compact and balanced hash codes in a Bayesian learning framework. The main idea is to reshape the standard cross-entropy loss for similarity-preserving learning such that it down-weighs the loss associated to highly-confident pairs. This idea leads to a novel priority cross-entropy loss, which prioritizes the training on uncertain pairs over confident pairs. Also, we propose another priority quantization loss, which prioritizes hard-to-quantize examples for generation of nearly lossless hash codes. Extensive experiments demonstrate that DPH can generate high-quality hash codes and yield state-of-the-art image retrieval results on three datasets, ImageNet, NUS-WIDE, and MS-COCO.

Deep hashing establishes efficient and effective image retrieval by end-to-end learning of deep representations and hash codes from similarity data. We present a compact coding solution, focusing on deep learning to quantization approach that has shown superior performance over hashing solutions for similarity retrieval. We propose Deep Triplet Quantization (DTQ), a novel approach to learning deep quantization models from the similarity triplets. To enable more effective triplet training, we design a new triplet selection approach, Group Hard, that randomly selects hard triplets in each image group. To generate compact binary codes, we further apply a triplet quantization with weak orthogonality during triplet training. The quantization loss reduces the codebook redundancy and enhances the quantizability of deep representations through back-propagation. Extensive experiments demonstrate that DTQ can generate high-quality and compact binary codes, which yields state-of-the-art image retrieval performance on three benchmark datasets, NUS-WIDE, CIFAR-10, and MS-COCO.

A technical challenge of deep learning is recognizing target classes without seen data. Zero-shot learning leverages semantic representations such as attributes or class prototypes to bridge source and target classes. Existing standard zero-shot learning methods may be prone to overfitting the seen data of source classes as they are blind to the semantic representations of target classes. In this paper, we study generalized zero-shot learning that assumes accessible to target classes for unseen data during training, and prediction on unseen data is made by searching on both source and target classes. We propose a novel Deep Calibration Network (DCN) approach towards this generalized zero-shot learning paradigm, which enables simultaneous calibration of deep networks on the confidence of source classes and uncertainty of target classes. Our approach maps visual features of images and semantic representations of class prototypes to a common embedding space such that the compatibility of seen data to both source and target classes are maximized. We show superior accuracy of our approach over the state of the art on benchmark datasets for generalized zero-shot learning, including AwA, CUB, SUN, and aPY.

Adversarial learning has been embedded into deep networks to learn disentangled and transferable representations for domain adaptation. Existing adversarial domain adaptation methods may struggle to align different domains of multimodal distributions that are native in classification problems. In this paper, we present conditional adversarial domain adaptation, a principled framework that conditions the adversarial adaptation models on discriminative information conveyed in the classifier predictions. Conditional domain adversarial networks (CDANs) are designed with two novel conditioning strategies: multilinear conditioning that captures the cross-covariance between feature representations and classifier predictions to improve the discriminability, and entropy conditioning that controls the uncertainty of classifier predictions to guarantee the transferability. Experiments testify that the proposed approach exceeds the state-of-the-art results on five benchmark datasets.

Hashing is widely applied to large-scale multimedia retrieval due to the storage and retrieval efficiency. Cross-modal hashing enables efficient retrieval of one modality from database relevant to a query of another modality. Existing work on cross-modal hashing assumes that heterogeneous relationship across modalities is available for learning to hash. This paper relaxes this strict assumption by only requiring heterogeneous relationship in some auxiliary dataset different from the query or database domain. We design a novel hybrid deep architecture, transitive hashing network (THN), to jointly learn cross-modal correlation from the auxiliary dataset, and align the data distributions of the auxiliary dataset with that of the query or database domain, which generates compact transitive hash codes for efficient cross-modal retrieval. Comprehensive empirical evidence validates that the proposed THN approach yields state of the art retrieval performance on standard multimedia benchmarks, i.e. NUS-WIDE and ImageNet-YahooQA.

Cross-modal similarity retrieval is a problem about designing a retrieval system that supports querying across content modalities, e.g., using an image to retrieve for texts. This paper presents a compact coding solution for efficient cross-modal retrieval, with a focus on the quantization approach which has already shown the superior performance over the hashing solutions in single-modal similarity retrieval. We propose a collective deep quantization (CDQ) approach, which is the first attempt to introduce quantization in end-to-end deep architecture for cross-modal retrieval. The major contribution lies in jointly learning deep representations and the quantizers for both modalities using carefully-crafted hybrid networks and well-specified loss functions. In addition, our approach simultaneously learns the common quantizer codebook for both modalities through which the cross-modal correlation can be substantially enhanced. CDQ enables efficient and effective cross-modal retrieval using inner product distance computed based on the common codebook with fast distance table lookup. Extensive experiments show that CDQ yields state of the art cross-modal retrieval results on standard benchmarks.

Compact coding has been widely applied to approximate nearest neighbor search for large-scale image retrieval, due to its computation efficiency and retrieval quality. This paper presents a compact coding solution with a focus on the deep learning to quantization approach, which improves retrieval quality by end-to-end representation learning and compact encoding and has already shown the superior performance over the hashing solutions for similarity retrieval. We propose Deep Visual-Semantic Quantization (DVSQ), which is the first approach to learning deep quantization models from labeled image data as well as the semantic information underlying general text domains. The main contribution lies in jointly learning deep visual-semantic embeddings and visual-semantic quantizers using carefully-designed hybrid networks and well-specified loss functions. DVSQ enables efficient and effective image retrieval by supporting maximum inner-product search, which is computed based on learned codebooks with fast distance table lookup. Comprehensive empirical evidence shows that DVSQ can generate compact binary codes and yield state-of-the-art similarity retrieval performance on standard benchmarks.

Two-stream convolutional networks have shown strong performance in video action recognition tasks. The key idea is to learn spatiotemporal features by fusing convolutional networks spatially and temporally. However, it remains unclear how to model the correlations between the spatial and temporal structures at multiple abstraction levels. First, the spatial stream tends to fail if two videos share similar backgrounds. Second, the temporal stream may be fooled if two actions resemble in short snippets, though appear to be distinct in the long term. We propose a novel spatiotemporal pyramid network to fuse the spatial and temporal features in a pyramid structure such that they can reinforce each other. From the architecture perspective, our network constitutes hierarchical fusion strategies which can be trained as a whole using a unified spatiotemporal loss. A series of ablation experiments support the importance of each fusion strategy. From the technical perspective, we introduce the spatiotemporal compact bilinear operator into video analysis tasks. This operator enables efficient training of bilinear fusion operations which can capture full interactions between the spatial and temporal features. Our final network achieves state-of-the-art results on standard video datasets.

Learning to hash has been widely applied to approximate nearest neighbor search for large-scale multimedia retrieval, due to its computation efficiency and retrieval quality. Deep learning to hash, which improves retrieval quality by end-to-end representation learning and hash encoding, has received increasing attention recently. Subject to the ill-posed gradient difficulty in the optimization with sign activations, existing deep learning to hash methods need to first learn continuous representations and then generate binary hash codes in a separated binarization step, which suffer from substantial loss of retrieval quality. This work presents HashNet, a novel deep architecture for deep learning to hash by continuation method with convergence guarantees, which learns exactly binary hash codes from imbalanced similarity data. The key idea is to attack the ill-posed gradient problem in optimizing deep networks with non-smooth binary activations by continuation method, in which we begin from learning an easier network with smoothed activation function and let it evolve during the training, until it eventually goes back to being the original, difficult to optimize, deep network with the sign activation function. Comprehensive empirical evidence shows that HashNet can generate exactly binary hash codes and yield state-of-the-art multimedia retrieval performance on standard benchmarks.

Deep networks have been successfully applied to learn transferable features for adapting models from a source domain to a different target domain. In this paper, we present joint adaptation networks (JAN), which learn a transfer network by aligning the joint distributions of multiple domain-specific layers across domains based on a joint maximum mean discrepancy (JMMD) criterion. Adversarial training strategy is adopted to maximize JMMD such that the distributions of the source and target domains are made more distinguishable. Learning can be performed by stochastic gradient descent with the gradients computed by back-propagation in linear-time. Experiments testify that our model yields state of the art results on standard datasets.

Deep networks trained on large-scale data can learn transferable features to promote learning multiple tasks. Since deep features eventually transition from general to specific along deep networks, a fundamental problem of multi-task learning is how to exploit the task relatedness underlying parameter tensors and improve feature transferability in the multiple task-specific layers. This paper presents Multilinear Relationship Networks (MRN) that discover the task relationships based on novel tensor normal priors over parameter tensors of multiple task-specific layers in deep convolutional networks. By jointly learning transferable features and multilinear relationships of tasks and features, MRN is able to alleviate the dilemma of negative-transfer in the feature layers and under-transfer in the classifier layer. Experiments show that MRN yields state-of-the-art results on three multi-task learning datasets.

The predictive learning of spatiotemporal sequences aims to generate future images by learning from the historical frames, where spatial appearances and temporal variations are two crucial structures. This paper models these structures by presenting a predictive recurrent neural network (PredRNN). This architecture is enlightened by the idea that spatiotemporal predictive learning should memorize both spatial appearances and temporal variations in a unified memory pool. Concretely, memory states are no longer constrained inside each LSTM unit. Instead, they are allowed to zigzag in two directions: across stacked RNN layers vertically and through all RNN states horizontally. The core of this network is a new Spatiotemporal LSTM (ST-LSTM) unit that extracts and memorizes spatial and temporal representations simultaneously. PredRNN achieves the state-of-the-art prediction performance on three video prediction datasets and is a more general framework, that can be easily extended to other predictive learning tasks by integrating with other architectures.

Hashing has been widely applied to approximate nearest neighbor search for large-scale multimedia retrieval. Supervised hashing improves the quality of hash coding by exploiting the semantic similarity on data pairs and has received increasing attention recently. For most existing supervised hashing methods for image retrieval, an image is first represented as a vector of hand-crafted or machine-learned features, then quantized by a separate quantization step that generates binary codes. However, suboptimal hash coding may be produced, since the quantization error is not statistically minimized and the feature representation is not optimally compatible with the hash coding. In this paper, we propose a novel Deep Quantization Network (DQN) architecture for supervised hashing, which learns image representation for hash coding and formally control the quantization error. The DQN model constitutes four key components: (1) a sub-network with multiple convolution-pooling layers to capture deep image representations; (2) a fully connected bottleneck layer to generate dimension-reduced representation optimal for hash coding; (3) a pairwise cosine loss layer for similarity-preserving learning; and (4) a product quantization loss for controlling hashing quality and the quantizability of bottleneck representation. Extensive experiments on standard image retrieval datasets show the proposed DQN model yields substantial boosts over latest state-of-the-art hashing methods.

Due to the storage and retrieval efficiency, hashing has been widely deployed to approximate nearest neighbor search for large-scale multimedia retrieval. Supervised hashing, which improves the quality of hash coding by exploiting the semantic similarity on data pairs, has received increasing attention recently. For most existing supervised hashing methods for image retrieval, an image is first represented as a vector of hand-crafted or machine-learned features, followed by another separate quantization step that generates binary codes. However, suboptimal hash coding may be produced, because the quantization error is not statistically minimized and the feature representation is not optimally compatible with the binary coding. In this paper, we propose a novel Deep Hashing Network (DHN) architecture for supervised hashing, in which we jointly learn good image representation tailored to hash coding and formally control the quantization error. The DHN model constitutes four key components: (1) a sub-network with multiple convolution-pooling layers to capture image representations; (2) a fully-connected hashing layer to generate compact binary hash codes; (3) a pairwise cross-entropy loss layer for similarity-preserving learning; and (4) a pairwise quantization loss for controlling hashing quality. Extensive experiments on standard image retrieval datasets show the proposed DHN model yields substantial boosts over latest state-of-the-art hashing methods.

Due to the storage and retrieval efficiency, hashing has been widely applied to approximate nearest neighbor search for large-scale multimedia retrieval. Cross-modal hashing, which enables efficient retrieval of images in response to text queries or vice versa, has received increasing attention recently. Most existing work on cross-modal hashing does not capture the spatial dependency of images and temporal dynamics of text sentences for learning powerful feature representations and cross-modal embeddings that mitigate the heterogeneity of different modalities. This paper presents a new Deep Visual-Semantic Hashing (DVSH) model that generates compact hash codes of images and sentences in an end-to-end deep learning architecture, which capture the intrinsic cross-modal correspondences between visual data and natural language. DVSH is a hybrid deep architecture that constitutes a visual-semantic fusion network for learning joint embedding space of images and text sentences, and two modality-specific hashing networks for learning hash functions to generate compact binary codes. Our architecture effectively unifies joint multimodal embedding and cross-modal hashing, which is based on a novel combination of Convolutional Neural Networks over images, Recurrent Neural Networks over sentences, and a structured max-margin objective that integrates all things together to enable learning of similarity-preserving and high-quality hash codes. Extensive empirical evidence shows that our DVSH approach yields state of the art results in cross-modal retrieval experiments on image-sentences datasets, i.e. standard IAPR TC-12 and large-scale Microsoft COCO.

Due to its storage and query efficiency, hashing has been widely applied to approximate nearest neighbor search from large-scale datasets. While there is increasing interest in cross-modal hashing which facilitates cross-media retrieval by embedding data from different modalities into a common Hamming space, how to distill the cross-modal correlation structure effectively remains a challenging problem. In this paper, we propose a novel supervised cross-modal hashing method, Correlation Autoencoder Hashing (CAH), to learn discriminative and compact binary codes based on deep autoencoders. Specifically, CAH jointly maximizes the feature correlation revealed by bimodal data and the semantic correlation conveyed in similarity labels, while embeds them into hash codes by nonlinear deep autoencoders. Extensive experiments clearly show the superior effectiveness and efficiency of CAH against the state-of-the-art hashing methods on standard cross-modal retrieval benchmarks.

The recent success of deep neural networks relies on massive amounts of labeled data. For a target task where labeled data is unavailable, domain adaptation can transfer a learner from a different source domain. In this paper, we propose a new approach to domain adaptation in deep networks that can jointly learn adaptive classifiers and transferable features from labeled data in the source domain and unlabeled data in the target domain. We relax a shared-classifier assumption made by previous methods and assume that the source classifier and target classifier differ by a residual function. We enable classifier adaptation by plugging several layers into deep network to explicitly learn the residual function with reference to the target classifier. We fuse features of multiple layers with tensor product and embed them into reproducing kernel Hilbert spaces to match distributions for feature adaptation. The adaptation can be achieved in most feed-forward models by extending them with new residual layers and loss functions, which can be trained efficiently via back-propagation. Empirical evidence shows that the new approach outperforms state of the art methods on standard domain adaptation benchmarks.

Efficient similarity retrieval from large-scale multimodal database is pervasive in modern search engines and social networks. To support queries across content modalities, the system should enable cross-modal correlation and computation-efficient indexing. While hashing methods have shown great potential in achieving this goal, current attempts generally fail to learn isomorphic hash codes in a seamless scheme, that is, they embed multiple modalities in a continuous isomorphic space and separately threshold embeddings into binary codes, which incurs substantial loss of retrieval accuracy. In this paper, we approach seamless multimodal hashing by proposing a novel Composite Correlation Quantization (CCQ) model. Specifically, CCQ jointly finds correlation-maximal mappings that transform different modalities into isomorphic latent space, and learns composite quantizers that convert the isomorphic latent features into compact binary codes. An optimization framework is devised to preserve both intra-modal similarity and inter-modal correlation through minimizing both reconstruction and quantization errors, which can be trained from both paired and partially paired data in linear time. A comprehensive set of experiments clearly show the superior effectiveness and efficiency of CCQ against the state of the art hashing methods for both unimodal and cross-modal retrieval.

Domain adaptation generalizes a learning model across source domain and target domain that are sampled from different distributions. It is widely applied to cross-domain data mining for reusing labeled information and mitigating labeling consumption. Recent studies reveal that deep neural networks can learn abstract feature representation, which can reduce, but not remove, the cross-domain discrepancy. To enhance the invariance of deep representation and make it more transferable across domains, we propose a unified deep adaptation framework for jointly learning transferable representation and classifier to enable scalable domain adaptation, by taking the advantages of both deep learning and optimal two-sample matching. The framework constitutes two inter-dependent paradigms, unsupervised pre-training for effective training of deep models using deep denoising autoencoders, and supervised fine-tuning for effective exploitation of discriminative information using deep neural networks, both learned by embedding the deep representations to reproducing kernel Hilbert spaces (RKHSs) and optimally matching different domain distributions. To enable scalable learning, we develop a linear-time algorithm using unbiased estimate that scales linearly to large samples. Extensive empirical results show that the proposed framework significantly outperforms state of the art methods on diverse adaptation tasks: sentiment polarity prediction, email spam filtering, newsgroup content categorization, and visual object recognition.

Recent studies reveal that a deep neural network can learn transferable features which generalize well to novel tasks for domain adaptation. However, as deep features eventually transition from general to specific along the network, the feature transferability drops significantly in higher layers with increasing domain discrepancy. Hence, it is important to formally reduce the dataset bias and enhance the transferability in task-specific layers. In this paper, we propose a new Deep Adaptation Network (DAN) architecture, which generalizes deep convolutional neural network to the domain adaptation scenario. In DAN, hidden representations of all task-specific layers are embedded in a reproducing kernel Hilbert space where the mean embeddings of different domain distributions can be explicitly matched. The domain discrepancy is further reduced using an optimal multi-kernel selection method for mean embedding matching. DAN can learn transferable features with statistical guarantees, and can scale linearly by unbiased estimate of kernel embedding. Extensive empirical evidence shows that the proposed architecture yields state-of-the-art image classification error rates on standard domain adaptation benchmarks.

Domain transfer learning generalizes a learning model across training data and testing data with different distributions. A general principle to tackle this problem is reducing the distribution difference between training data and testing data such that the generalization error can be bounded. Current methods typically model the sample distributions in input feature space, which depends on nonlinear feature mapping to embody the distribution discrepancy. However, this nonlinear feature space may not be optimal for the kernel-based learning machines. To this end, we propose a transfer kernel learning (TKL) approach to learn a domain-invariant kernel by directly matching source and target distributions in the reproducing kernel Hilbert space (RKHS). Specifically, we design a family of spectral kernels by extrapolating target eigensystem on source samples with Mercer's theorem. The spectral kernel minimizing the approximation error to the ground truth kernel is selected to construct domain-invariant kernel machines. Comprehensive experimental evidence on a large number of text categorization, image classification, and video event recognition datasets verifies the effectiveness and efficiency of the proposed TKL approach over several state-of-the-art methods.

Visual domain adaptation, which learns an accurate classifier for a new domain using labeled images from an old domain, has shown promising value in computer vision yet still been a challenging problem. Most prior works have explored two learning strategies independently for domain adaptation: feature matching and instance reweighting. In this paper, we show that both strategies are important and inevitable when the domain difference is substantially large. We therefore put forward a novel Transfer Joint Matching (TJM) approach to model them in a unified optimization problem. Specifically, TJM aims to reduce the domain difference by jointly matching the features and reweighting the instances across domains in a principled dimensionality reduction procedure, and construct new feature representation that is invariant to both the distribution difference and the irrelevant instances. Comprehensive experimental results verify that TJM can significantly outperform competitive methods for cross-domain image recognition problems.

Sparse coding has received considerable research attentions due to its competitive performance for SPM-based image classification algorithms. In sparse coding, each low-level image descriptor (e.g., SIFT) is quantized into a sparse vector using an over-complete dictionary. Two typical schemes for achieving the code sparsity are imposing ℓ1-sparsity penalty on the coding coefficients, or selecting a set of fc-nearest-neighbor bases from the dictionary for locality-aware encoding. In this paper, we discover that different coding schemes usually produce substantially inconsistent coefficients, each preferring either ℓ1-sparsity or bases-locality. We therefore conjecture that different schemes should be explored simultaneously to further enhance the quantization quality. To this end, we propose a novel ensemble framework, Local Hybrid Coding (LHC), to formalize a unified optimization problem for different coding schemes. Specifically, we quantize each image descriptor using two disjoint sets of dictionaries, fcNN bases and non-fcNN bases, from which we efficiently compute a hybrid representation comprising of local coding and sparse coding, respectively. Extensive experiments on three benchmarks verify that LHC can remarkably outperform several state-of-the-art methods for image classification tasks, and bare comparable complexity to the most efficient coding methods.

Transfer learning is established as an effective technology to leverage rich labeled data from some source domain to build an accurate classifier for the target domain. The basic assumption is that the input domains may share certain knowledge structure, which can be encoded into common latent factors and extracted by preserving important property of original data, e.g., statistical property and geometric structure. In this paper, we show that different properties of input data can be complementary to each other and exploring them simultaneously can make the learning model robust to the domain difference. We propose a general framework, referred to as Graph Co-Regularized Transfer Learning (GTL), where various matrix factorization models can be incorporated. Specifically, GTL aims to extract common latent factors for knowledge transfer by preserving the statistical property across domains, and simultaneously, refine the latent factors to alleviate negative transfer by preserving the geometric structure in each domain. Based on the framework, we propose two novel methods using NMF and NMTF, respectively. Extensive experiments verify that GTL can significantly outperform state-of-the-art learning methods on several public text and image datasets.

Domain transfer learning, which learns a target classifier using labeled data from a different distribution, has shown promising value in knowledge discovery yet still been a challenging problem. Most previous works designed adaptive classifiers by exploring two learning strategies independently: distribution adaptation and label propagation. In this paper, we propose a novel transfer learning framework, referred to as Adaptation Regularization based Transfer Learning (ARTL), to model them in a unified way based on the structural risk minimization principle and the regularization theory. Specifically, ARTL learns the adaptive classifier by simultaneously optimizing the structural risk functional, the joint distribution matching between domains, and the manifold consistency underlying marginal distribution. Based on the framework, we propose two novel methods using Regularized Least Squares (RLS) and Support Vector Machines (SVMs), respectively, and use the Representer theorem in reproducing kernel Hilbert space to derive corresponding solutions. Comprehensive experiments verify that ARTL can significantly outperform state-of-the-art learning methods on several public text and image datasets.

Sparse coding learns a set of basis functions such that each input signal can be well approximated by a linear combination of just a few of the bases. It has attracted increasing interest due to its state-of-the-art performance in BoW based image representation. However, when labeled and unlabeled images are sampled from different distributions, they may be quantized into different visual words of the codebook and encoded with different representations, which may severely degrade classification performance. In this paper, we propose a Transfer Sparse Coding (TSC) approach to construct robust sparse representations for classifying cross-distribution images accurately. Specifically, we aim to minimize the distribution divergence between the labeled and unlabeled images, and incorporate this criterion into the objective function of sparse coding to make the new representations robust to the distribution difference. Experiments show that TSC can significantly outperform state-of-the-art methods on three types of computer vision datasets.

Transfer learning is established as an effective technology in computer vision for leveraging rich labeled data in the source domain to build an accurate classifier for the target domain. However, most prior methods have not simultaneously reduced the difference in both the marginal distribution and conditional distribution between domains. In this paper, we put forward a novel transfer learning approach, referred to as Joint Distribution Adaptation (JDA). Specifically, JDA aims to jointly adapt both the marginal distribution and conditional distribution in a principled dimensionality reduction procedure, and construct new feature representation that is effective and robust for substantial distribution difference. Extensive experiments verify that JDA can significantly outperform several state-of-the-art methods on four types of cross-domain image classification problems.

Cross-domain text classification aims to automatically train a precise text classifier for a target domain by using labeled text data from a related source domain. To this end, the distribution gap between different domains has to be reduced. In previous works, a certain number of shared latent features (e.g., latent topics, principal components, etc.) are extracted to represent documents from different domains, and thus reduce the distribution gap. However, only relying the shared latent features as the domain bridge may limit the amount of knowledge transferred. This limitation is more serious when the distribution gap is so large that only a small number of latent features can be shared between domains. In this paper, we propose a novel approach named Topic Correlation Analysis (TCA), which extracts both the shared and the domain-specific latent features to facilitate effective knowledge transfer. In TCA, all word features are first grouped into the shared and the domain-specific topics using a joint mixture model. Then the correlations between the two kinds of topics are inferred and used to induce a mapping between the domain-specific topics from different domains. Finally, both the shared and the mapped domain-specific topics are utilized to span a new shared feature space where the supervised knowledge can be effectively transferred. The experimental results on two real-world data sets justify the superiority of the proposed method over the stat-of-the-art baselines.

Transfer learning proves to be effective for leveraging labeled data in the source domain to build an accurate classifier in the target domain. The basic assumption behind transfer learning is that the involved domains share some common latent factors. Previous methods usually explore these latent factors by optimizing two separate objective functions, i.e., either maximizing the empirical likelihood, or preserving the geometric structure. Actually, these two objective functions are complementary to each other and optimizing them simultaneously can make the solution smoother and further improve the accuracy of the final model. In this paper, we propose a novel approach called Graph co-regularized Transfer Learning (GTL) for this purpose, which integrates the two objective functions seamlessly into one unified optimization problem. Thereafter, we present an iterative algorithm for the optimization problem with rigorous analysis on convergence and complexity. Our empirical study on two open data sets validates that GTL can consistently improve the classification accuracy compared to the state-of-the-art transfer learning methods.

Transfer learning aims to leverage the knowledge in the source domain to facilitate the learning tasks in the target domain. It has attracted extensive research interests recently due to its effectiveness in a wide range of applications. The general idea of the existing methods is to utilize the common latent structure shared across domains as the bridge for knowledge transfer. These methods usually model the common latent structure by using either the marginal distribution or the conditional distribution. However, without exploring the duality between these two distributions, these single bridge methods may not achieve optimal capability of knowledge transfer. In this paper, we propose a novel approach, Dual Transfer Learning (DTL), which simultaneously learns the marginal and conditional distributions, and exploits the duality between them in a principled way. The key idea behind DTL is that learning one distribution can help to learn the other. This duality property leads to mutual reinforcement when adapting both distributions across domains to transfer knowledge. The proposed method is formulated as an optimization problem based on joint nonnegative matrix trifactorizations (NMTF). The two distributions are learned from the decomposed latent factors that exhibit the duality property. An efficient alternating minimization algorithm is developed to solve the optimization problem with convergence guarantee. Extensive experimental results demonstrate that DTL is more effective than alternative transfer learning methods.

Transfer learning targets to leverage knowledge from one domain for tasks in a new domain. It finds abundant applications, such as text/sentiment classification. Many previous works are based on cluster analysis, which assume some common clusters shared by both domains. They mainly focus on the one-to-one cluster correspondence to bridge different domains. However, such a correspondence scheme might be too strong for real applications where each cluster in one domain corresponds to many clusters in the other domain. In this paper, we propose a Cluster Correspondence Inference (CCI) method to iteratively infer many-to-many correspondence among clusters from different domains. Specifically, word clusters and document clusters are exploited for each domain using nonnegative matrix factorization, then the word clusters from different domains are corresponded in a many-to-many scheme, with the help of shared word space as a bridge. These two steps are run iteratively and label information is transferred from source domain to target domain through the inferred cluster correspondence. Experiments on various real data sets demonstrate that our method outperforms several state-of-the-art approaches for cross-domain text classification.