In this paper, we investigate code-integrated reasoning (CIR), where models generate code when necessary and integrate feedback by executing it through a code interpreter. To acquire this capability, models must learn when and how to use external code tools effectively, which is supported by tool-augmented reinforcement learning (RL). Despite its benefits, tool-augmented RL can still suffer from potential instability in the learning dynamics. In light of this challenge, we present a systematic approach ETIR (Effective TIR) to improving the training effectiveness and stability of tool-augmented RL for code-integrated reasoning. Specifically, we develop enhanced training strategies that balance exploration and stability, progressively building tool-use capabilities while improving reasoning performance. Through extensive experiments on five mainstream mathematical reasoning benchmarks, our model demonstrates significant performance improvements over multiple competitive baselines. Furthermore, we conduct an in-depth analysis of the mechanism of code-integrated reasoning, revealing several key insights, such as the extension of model’s capability boundaries and the simultaneous improvement of reasoning efficiency through code integration. These findings underscore the potential of code-integrated reasoning as a scalable paradigm for advancing robust and efficient language model reasoning.

The generative mechanisms behind real-world event sequences are often heterogeneous, leading to data that possesses inherent clustering structures. However, most existing temporal point processes (TPPs) treat different event sequences independently, without leveraging the clustering structures when predicting events. In this study, we design and learn a novel semi-transductive temporal point process (ST-TPP), which explicitly improves prediction performance by co-training sequence clusters. In particular, given a set of event sequences, our method learns a neural TPP together with cluster centers of the sequences. Besides maximizing the likelihood of the event sequences, we leverage a data-based kernel matrix and prior knowledge to regularize the sequence embeddings, leading to a Gromov-Wasserstein barycentric (GWB) regularizer. Based on the optimal transport plans associated with the GWB regularizer, we derive the cluster centers by the push-forward of the sequence embeddings. When a new sequence comes, the learned model first assigns a cluster center to the sequence and then jointly encodes the sequence and the cluster center to predict future events, leading to a semi-transductive prediction scheme. Experiments demonstrate that ST-TPP achieves competitive sequence clustering results and strong prediction performance.

Medication recommendation aims to generate safe and effective medication sets from health records. However, accurately recommending medications hinges on inferring a patient’s latent clinical condition from sparse and noisy observations, which requires both (i) preserving the visit-level combinatorial semantics of co-occurring diagnoses/procedures and (ii) leveraging informative historical references through effective, visit-conditioned retrieval. Most existing methods fall short in one of these aspects: graph-based modeling often fragments higher-order intra-visit patterns into pairwise relations, while inter-visit augmentation methods commonly exhibit an imbalance between learning a globally stable representation space and performing dynamic retrieval within it. To address these limitations, this article proposes HypeMed , a two-stage hypergraph-based framework unifying intra-visit coherence modeling and inter-visit augmentation. HypeMed consists of two components: MedRep for representation pretraining and SimMR for similarity-enhanced recommendation. In the first stage, MedRep encodes clinical visits as hyperedges via knowledge-aware contrastive pretraining, creating a globally consistent, retrieval-friendly embedding space. In the second stage, SimMR performs dynamic retrieval within this space, fusing retrieved references with the patient’s longitudinal data to refine medication prediction. Evaluation on real-world benchmarks shows that HypeMed outperforms state-of-the-art baselines in both recommendation precision and DDI reduction, simultaneously enhancing the effectiveness and safety of clinical decision support. The implementation is publicly available at https://github.com/xansar/HypeMed .

As a fundamental problem of graph analysis, graph visualization aims to embed a set of graphs in a low-dimensional (e.g., 2D) space and provide insights into their distribution and clustering structure. Focusing on this problem, we propose a novel Wasserstein t-distributed embedding (WatE) method, leading to an information-enriched graph visualization paradigm. Our method learns a graph neural network to represent each graph as the mean and covariance of its node embedding distribution. Accordingly, our method can visualize each graph as an ellipse (determined by the mean and the covariance) rather than a single point. The positions of different ellipses reveal the relations among different graphs as traditional visualization methods do, while the size and shape of an ellipse preserve the node-level structural information of the corresponding graph. We propose a regularized t-distributed stochastic neighbor embedding (Rt-SNE) framework to learn the visualization model, deriving a Wasserstein distance-based Student's t-distribution of graph pairs and fitting the distribution to the data distribution under regularization. Both subjective and objective evaluations demonstrate that WatE achieves encouraging performance in various graph visualization and clustering tasks.

Multi-modal learning aims to learn predictive models based on the data from different modalities. However, due to the requirement of data security and privacy protection, real-world multi-modal data are often scattered to different agents and cannot be shared across the agents, which limits the application of existing multi-modal learning methods. To achieve robust multi-modal learning in such a challenging scenario, we propose a novel optimal transport-based mixer (OTM), which works as an effective latent code alignment and augmentation method for unaligned and distributed multi-modal data. In particular, we train a Wasserstein autoencoder (WAE) for each agent, which encodes its single modal samples in a latent space. Through a central server, the proposed OTM computes a stochastic fused Gromov-Wasserstein barycenter (FGWB) to mix different modalities' latent codes, so that each agent applies the barycenter to reconstruct its samples. This method neither requires well-aligned multi-modal data nor assumes the data to share the same latent distribution, and each agent can learn a specific model based on multi-modal data while achieving inference based on its local modality. Experiments on multi-modal clustering and classification demonstrate that the models learned with the OTM method outperform the corresponding baselines.

An event sequence generated by a temporal point process is often associated with a hidden and structured event branching process that captures the triggering relations between its historical and current events. In this study, we design a new plug-and-play module based on the Bregman ADMM (BADMM) algorithm, which infers event branches associated with event sequences in the maximum likelihood estimation framework of temporal point processes (TPPs). Specifically, we formulate the inference of event branches as an optimization problem of event transition matrix under sparse and low-rank constraints, which is embedded in existing TPP models or their learning paradigms. We can implement this optimization problem based on subspace clustering and sparse group-lasso, respectively, and solve it using the Bregman ADMM algorithm, whose unrolling leads to the proposed BADMM module. When learning a classic TPP (e.g., Hawkes process) by the expectation-maximization algorithm, the BADMM module helps derive structured responsibility matrices in the E-step. Similarly, the BADMM module helps derive low-rank and sparse attention maps for the neural TPPs with self-attention layers. The structured responsibility matrices and attention maps, which work as learned event transition matrices, indicate event branches, e.g., inferring isolated events and those key events triggering many subsequent events. Experiments on both synthetic and real-world data show that plugging our BADMM module into existing TPP models and learning paradigms can improve model performance and provide us with interpretable structured event branches.

This paper addresses a promising yet underexplored task, Image-to-Sounding-Video (I2SV) generation, which animates a static image and generates synchronized sound simultaneously. Despite advances in video and audio generation models, challenges remain to develop a unified model for generating naturally sounding videos. In this work, we propose a novel approach that leverages two separate pretrained diffusion models and makes vision and audio influence each other during generation based on the Diffusion Transformer (DiT) architecture. First, the individual video and audio pretrained generation models are decomposed into input, output, and expert sub-modules. We propose using a unified joint DiT block to integrate the expert sub-modules to effectively model the interaction between the two modalities, resulting in high-quality I2SV generation. Then, we introduce a joint classifier-free guidance technique to boost the performance during joint generation. Finally, we conduct extensive experiments on three popular benchmark datasets, and in both objective and subjective evaluation our method surpass all the baseline methods in almost all metrics. Case studies show our generated sounding videos are high quality and synchronized between video and audio.

As a very common type of video, face videos often appear in movies, talk shows, live broadcasts, and other scenes. Real-World online videos are often plagued by degradations such as blurring and quantization noise, due to the high compression ratio caused by high communication costs and limited transmission bandwidth. These degradations have a particularly serious impact on face videos because the human visual system is highly sensitive to facial details. Despite the significant advancement in video face enhancement, current methods still suffer from i) long processing time and ii) inconsistent spatial-temporal visual effects (e.g., flickering). This study proposes a novel and efficient blind video face enhancement method to overcome the above two challenges, restoring high-quality videos from their compressed low-quality versions with an effective de-flickering mechanism. In particular, the proposed method develops upon a 3D-VQGAN backbone associated with spatial-temporal codebooks recording high-quality portrait features and residual-based temporal information. We develop a two-stage learning framework for the model. In Stage I, we learn the model with a regularizer mitigating the codebook collapse problem. In Stage II, we learn two transformers to look up code from the codebooks and further update the encoder of low-quality videos. Experiments conducted on the VFHQ-Test dataset demonstrate that our method surpasses the current state-of-the-art blind face video restoration and de-flickering methods on both efficiency and effectiveness. Code is available at https://github.com/Dixin-Lab/BFVR-STC.

Domain adaptation aims to align the data scattered in different domains, which is important for developing generalizable machine learning models. However, real-world data in different domains are often heterogeneous, requiring alignment at both sample and feature levels. In this study, we develop a new optimal transport-based method, unbalanced co-relational optimal transport (UCROT), to achieve robust heterogeneous data alignment. Given the data in different domains, we define a co-relational optimal transport problem, jointly inferring the optimal transport (OT) plans defined at the sample and feature levels. The OT plans indicate the sample correspondence and feature correlation across different domains. In addition, we relax the doubly stochastic constraints of the OT plans to the KL-divergence regularization of their marginals, which enhances the robustness of our method to the sample- and feature-level outliers and leads to the proposed UCROT method. Experiments on the heterogeneous domain adaptation and co-clustering tasks demonstrate the superiority of UCROT.

Polymer conformation generation is a critical task that enables atomic-level studies of diverse polymer materials. While significant advances have been made in designing conformation generation methods for small molecules and proteins, these methods struggle to generate polymer conformations due to their unique structural characteristics. Meanwhile, the scarcity of polymer conformation datasets further limits progress, making this important area largely unexplored. In this work, we propose PolyConf, a pioneering tailored polymer conformation generation method that leverages hierarchical generative models to unlock new possibilities. Specifically, we decompose the polymer conformation into a series of local conformations (i.e., the conformations of its repeating units), generating these local conformations through an autoregressive model, and then generating their orientation transformations via a diffusion model to assemble them into the complete polymer conformation. Moreover, we develop the first benchmark with a high-quality polymer conformation dataset derived from molecular dynamics simulations to boost related research in this area. The comprehensive evaluation demonstrates that PolyConf consistently outperforms existing conformation generation methods, thus facilitating advancements in polymer modeling and simulation. The whole work is available at https://polyconf-icml25.github.io.

Predicting molecular ground-state conformation (i.e., energy-minimized conformation) is crucial for many chemical applications such as molecular docking and property prediction. Classic energy-based simulation is time-consuming when solving this problem, while existing learning-based methods have advantages in computational efficiency but sacrifice accuracy and interpretability. In this work, we propose a novel and effective method to bridge the energy-based simulation and the learning-based strategy, which designs and learns a Wasserstein gradient flow-driven SE(3)-Transformer, called WGFormer, for ground-state conformation prediction. Specifically, our method tackles this task within an auto-encoding framework, which encodes low-quality conformations by the proposed WGFormer and decodes corresponding ground-state conformations by an MLP. The architecture of WGFormer corresponds to Wasserstein gradient flows --- it optimizes conformations by minimizing an energy function defined on the latent mixture models of atoms, thereby significantly improving performance and interpretability. Extensive experiments demonstrate that our method consistently outperforms state-of-the-art competitors, providing a new and insightful paradigm to predict ground-state conformation. The code is available at https://github.com/FanmengWang/WGFormer.

Protein backbone generation plays a central role in de novo protein design and is significant for many biological and medical applications. Although diffusion and flow-based generative models provide potential solutions to this challenging task, they often generate proteins with undesired designability and suffer computational inefficiency. In this study, we propose a novel rectified quaternion flow (ReQFlow) matching method for fast and high-quality protein backbone generation. In particular, our method generates a local translation and a 3D rotation from random noise for each residue in a protein chain, which represents each 3D rotation as a unit quaternion and constructs its flow by spherical linear interpolation (SLERP) in an exponential format. We train the model by quaternion flow (QFlow) matching with guaranteed numerical stability and rectify the QFlow model to accelerate its inference and improve the designability of generated protein backbones, leading to the proposed ReQFlow model. Experiments show that ReQFlow achieves on-par performance in protein backbone generation while requiring much fewer sampling steps and significantly less inference time (e.g., being 37$\times$ faster than RFDiffusion and 63$\times$ faster than Genie2 when generating a backbone of length 300), demonstrating its effectiveness and efficiency.

As an essential movie promotional tool, trailers are designed to capture the audience's interest through the skillful editing of key movie shots. Although some attempts have been made for automatic trailer generation, existing methods often rely on predefined rules or manual fine-grained annotations and fail to fully leverage the multi-modal information of movies, resulting in unsatisfactory trailer generation results. In this study, we introduce a weakly-supervised trailer generation method driven by multi-modal semantic consistency. Specifically, we design a multi-modal trailer generation framework that selects and sorts key movie shots based on input music and movie metadata (e.g., category tags and plot keywords) and adds narration to the generated trailer based on movie subtitles. We utilize two pseudo-scores derived from the proposed framework as labels and thus train the model under a weakly-supervised learning paradigm, ensuring trailerness consistency for key shot selection and emotion consistency for key shot sorting, respectively. As a result, we can learn the proposed model solely based on movie-trailer pairs without any fine-grained annotations. Both objective experimental results and subjective user studies demonstrate the superior performance of our method over previous works. The code is available at https://github.com/Dixin-Lab/MMSC.

Session-based recommendation aims to predict the next item in the absence of user profiles and extensive interactions. Recent recommender techniques, such as attention mechanisms and graph neural networks, have shown promising results. However, most existing methods treat each item within a session as an independent intent unit, thus focusing solely on single-level user interests. This strategy overlooks the presence of multiple granularity levels of user intent that may exist within a session. To address this, we bundle locally consecutive items together as session snippets, which are modeled as high-level user intent units in the form of set-based and sequence-based representations. The user intent units are further enriched by aggregating multi-granularity neighbor information through two types of representations in the Heterogeneous Session Graph. We then treat the last intent unit from each guanularity as a query vector, learning the corresponding session representation using an attention mechanism. Based on this, we calculate a separate recommendation result. Next, we fine-tune the pre-trained meta-weight network to customize the weights for the recommendation results at different granularities. These weights are then used to fuse the results and generate the final recommendation. Extensive experiments on three real-world datasets demonstrate that our approach outperforms state-of-the-art methods.

Data-centric methods have shown great potential in understanding and predicting spatiotemporal dynamics, enabling better design and control of the object system. However, deep learning models often lack interpretability, fail to obey intrinsic physics, and struggle to cope with the various domains. While geometry-based methods, e.g., graph neural networks (GNNs), have been proposed to further tackle these challenges, they still need to find the implicit physical laws from large datasets and rely excessively on rich labeled data. In this paper, we herein introduce the conservation-informed GNN (CiGNN), an end-to-end explainable learning framework, to learn spatiotemporal dynamics based on limited training data. The network is designed to conform to the general conservation law via symmetry, where conservative and non-conservative information passes over a multiscale space enhanced by a latent temporal marching strategy. The efficacy of our model has been verified in various spatiotemporal systems based on synthetic and real-world datasets, showing superiority over baseline models. Results demonstrate that CiGNN exhibits remarkable accuracy and generalizability, and is readily applicable to learning for prediction of various spatiotemporal dynamics in a spatial domain with complex geometry.

Multimodal recommender systems exploit multi-modal information of items as an auxiliary, to learn more informative representations and achieve better performances. Most existing methods directly combine ID representations and different modality representations in trivial ways (e.g., element-wise sum, or attention mechanism). However, these strategies hardly accurately align different representations of the same user/item (e.g., the ID and modality representations), since the ID representation carries the overall semantic features of the user/item while each modality representation merely contains its partial semantic content, and the semantic information from different modalities may be partially complementary. To address this issue, we design a disentangled representation learning to decouple different representations of the user/item into different semantic components, in which any two components should carry either complementary or common semantic information. In each semantic scene, we first calculate the interaction probability between each pairwise components (e.g., a user component and an item component), and then customize their weights with a meta-weight net to integrate these probabilities as the current semantic score. Afterward, we fuse different semantic scores to estimate the final probability of user-item interaction. In addition, we introduce two contrastive learning tasks to maintain the consistency between multiple self-supervised views. Experimental evaluations on three datasets indicate that the proposed approach consistently outperforms current state-of-the-art baselines.

Owing to the incorporation of multimodal content, multimodal recommender systems contribute to powerful representation learning and promote personalized recommendations. However, most existing methods directly mix various modality embeddings and ID embeddings by element-wise sum as final representations. This strategy hardly distinguishes the importance of different representations and entangles the ID and modality effects, and thus is unable to fully leverage multimodal features to facilitate personalized recommendations. To address this problem, we design a meta-weight net to learn the relevance between pairwise embeddings (including modality embeddings and ID embeddings) from users and items, respectively. Specifically, the proposed model first performs graph convolutional networks to generate the modality and ID representations for users and items. For each type of user representation, we calculate its relevance weights with all types of item representations with the designed meta-weight net. Afterward, we utilize these relevance weights to adaptively integrate the corresponding probability scores to calculate the final recommendation result. Moreover, we adopt contrastive learning to alleviate the data sparsity issue and introduce the KL divergence constraint to better align user preferences and multimodal content. Experiments on three public datasets demonstrate that the proposed approach performs better than the state-of-the-art methods.

Adapting large-scale foundation models in multi-task scenarios often suffers from task conflict and oblivion. To mitigate such issues, we propose a novel "model MoE-ization" strategy that leads to a conflict- and oblivion-resistant multi-task adaptation method. Given a weight matrix of a pre-trained model, our method applies SVD to it and introduces a learnable router to adjust its singular values based on tasks and samples. Accordingly, the weight matrix becomes a Mixture of Orthogonal Rank-one Experts (MoORE), in which each expert corresponds to the outer product of a left singular vector and the corresponding right one. We can improve the model capacity by imposing a learnable orthogonal transform on the right singular vectors. Unlike low-rank adaptation (LoRA) and its MoE-driven variants, MoORE guarantees the experts' orthogonality and maintains the column space of the original weight matrix. These two properties make the adapted model resistant to the conflicts among the new tasks and the oblivion of its original tasks, respectively. Experiments on various datasets demonstrate that MoORE outperforms existing multi-task adaptation methods consistently, showing its superiority in terms of conflict- and oblivion-resistance. The code is available at https://github.com/DaShenZi721/MoORE.

Abstract Geometric graphs are a special kind of graph with geometric features, which are vital to model many scientific problems. Unlike generic graphs, geometric graphs often exhibit physical symmetries of translations, rotations, and reflections, making them ineffectively processed by current Graph Neural Networks (GNNs). To address this issue, researchers proposed a variety of geometric GNNs equipped with invariant/equivariant properties to better characterize the geometry and topology of geometric graphs. Given the current progress in this field, it is imperative to conduct a comprehensive survey of data structures, models, and applications related to geometric GNNs. In this paper, based on the necessary but concise mathematical preliminaries, we formalize geometric graph as the data structure, on top of which we provide a unified view of existing models from the geometric message passing perspective. Additionally, we summarize the applications as well as the related datasets to facilitate later research for methodology development and experimental evaluation. We also discuss the challenges and future potential directions of geometric GNNs at the end of this survey.

Graph Neural Networks (GNNs) have demonstrated remarkable effectiveness in recommendation systems. However, conventional graph-based recommenders, such as LightGCN, require maintaining embeddings of size d for each node, resulting in a parameter complexity of O ( n X d ), where n represents the total number of users and items. This scaling pattern poses significant challenges for deployment on large-scale graphs encountered in real-world applications. To address this scalability limitation, we propose Lighter-X , an efficient and modular framework that can be seamlessly integrated with existing GNN-based recommender architectures. Our approach substantially reduces both parameter size and computational complexity while preserving the theoretical guarantees and empirical performance of the base models, thereby enabling practical deployment at scale. Specifically, we analyze the original structure and inherent redundancy in their parameters, identifying opportunities for optimization. Based on this insight, we propose an efficient compression scheme for the sparse adjacency structure and high-dimensional embedding matrices, achieving a parameter complexity of O ( h X d ), where h >> n. Furthermore, the model is optimized through a decoupled framework, reducing computational complexity during the training process and enhancing scalability. Extensive experiments demonstrate that Lighter-X achieves comparable performance to baseline models with significantly fewer parameters. In particular, on large-scale interaction graphs with millions of edges, we are able to attain even better results with only 1% of the parameter over LightGCN.

Hyperbolic spaces have been considered pervasively for embedding hierarchically structured data in the recent decade. However, there is a lack of studies focusing on efficient distance metrics for comparing probability distributions in hyperbolic spaces. To bridge the gap, we propose a novel metric called the hyperbolic space-filling curve projection Wasserstein (SFW) distance. The idea is to first project two probability distributions onto a space-filling curve to obtain a closed-form coupling between them and then calculate the transport distance between these two distributions in the hyperbolic space accordingly. Theoretically, we show the SFW distance is a proper metric and is well-defined for probability measures with bounded supports. Statistical convergence rates for the proposed estimator are provided as well. Moreover, we propose two variants of the SFW distance based on geodesic and horospherical projections, respectively, to combat the curse-of-dimensionality. Empirical results on synthetic and real-world data indicate that the SFW distance can effectively serve as a surrogate of the popular Wasserstein distance with low complexity.

Graph neural networks have been widely used to represent 3D molecules, which capture molecular attributes and geometric information through various message-passing mechanisms. This study proposes a novel quaternion message-passing (QMP) module that can be plugged into many existing 3D molecular representation models and enhance their power for distinguishing molecular conformations. In particular, our QMP module represents the 3D rotations between one chemical bond and its neighbor bonds as a quaternion sequence. Then, it aggregates the rotations by the chained Hamilton product of the quaternions. The real part of the output quaternion is invariant to the global 3D rotations of molecules but sensitive to the local torsions caused by twisting bonds, providing discriminative information for training molecular conformation representation models. In theory, we prove that considering these features enables invariant GNNs to distinguish the conformations caused by bond torsions. We encapsulate the QMP module with acceleration, so combining existing models with the QMP requires merely one-line code and little computational cost. Experiments on various molecular datasets show that plugging our QMP module into existing invariant GNNs leads to consistent and significant improvements in molecular conformation representation and downstream tasks.

With the increasing application of deep learning to solve scientific problems in biochemistry, molecular federated learning has become popular due to its ability to offer distributed privacy-preserving solutions. However, most existing molecular federated learning methods rely on joint training with public datasets, which are difficult to obtain in practice. These methods also fail to leverage multi-modal molecular representations effectively. To address the above issues, we propose a novel framework, Federated Heterogeneous Contrastive Distillation (FedHCD), which enables to jointly train global models from clients with heterogeneous data modalities, learning tasks, and molecular models. To aggregate data representations of different modalities in a data-free manner, we design a global multi-modal contrastive strategy to align the representation of clients without public dataset. Utilizing intrinsic characteristics of molecular data in different modalities, we tackle the exacerbation of local model drift and data Non-IIDness caused by multi-modal clients. We introduce a multi-view contrastive knowledge transfer to extract features from atoms, substructures, and molecules, solving the issue of information distillation failure due to dimensional biases in different data modalities. Our evaluations on eight real-world molecular datasets and ablation experiments show that FedHCD outperforms other state-of-the-art FL methods, irrespective of whether or not they use public datasets.

Polymers are high-molecular-weight compounds constructed by the covalent bonding of numerous identical or similar monomers so that their 3D structures are complex yet exhibit unignorable regularity. Typically, the properties of a polymer, such as plasticity, conductivity, bio-compatibility, and so on, are highly correlated with its 3D structure. However, existing polymer property prediction methods heavily rely on the information learned from polymer SMILES sequences (P-SMILES strings) while ignoring crucial 3D structural information, resulting in sub-optimal performance. In this work, we propose MMPolymer, a novel multimodal multitask pretraining framework incorporating polymer 1D sequential and 3D structural information to encourage downstream polymer property prediction tasks. Besides, considering the scarcity of polymer 3D data, we further introduce the "Star Substitution" strategy to extract 3D structural information effectively. During pretraining, in addition to predicting masked tokens and recovering clear 3D coordinates, MMPolymer achieves the cross-modal alignment of latent representations. Then we further fine-tune the pretrained MMPolymer for downstream polymer property prediction tasks in the supervised learning paradigm. Experiments show that MMPolymer achieves state-of-the-art performance in downstream property prediction tasks. Moreover, given the pretrained MMPolymer, utilizing merely a single modality in the fine-tuning phase can also outperform existing methods, showcasing the exceptional capability of MMPolymer in polymer feature extraction and utilization.

As a significant step for human face modeling, editing, and generation, face landmarking aims at extracting facial keypoints from images. A generalizable face landmarker is required in practice because real-world facial images, e.g., the avatars in animations and games, are often stylized in various ways. However, achieving generalizable face land-marking is challenging due to the diversity of facial styles and the scarcity of labeled stylized faces. In this study, we propose a simple but effective paradigm to learn a generalizable face landmarker based on labeled real human faces and unlabeled stylized faces. Our method learns the face landmarker as the key module of a conditional face warper. Given a pair of real and stylized facial images, the conditional face warper predicts a warping field from the real face to the stylized one, in which the face landmarker predicts the ending points of the warping field and provides us with high-quality pseudo landmarks for the corresponding stylized facial images. Applying an alternating optimization strategy, we learn the face landmarker to minimize i) the discrepancy between the stylized faces and the warped real ones and ii) the prediction errors of both real and pseudo landmarks. Experiments on various datasets show that our method outperforms existing state-of-the-art domain adaptation methods in face landmarking tasks, leading to a face landmarker with better generalizability. Code is available at https://plustwo0.github.io/project-face-landmarker.

Many existing anomaly detection methods assume the availability of a large-scale normal dataset. But for many applications, limited by resources, removing all anomalous samples from a large un-labeled dataset is unrealistic, resulting in contaminated datasets. To detect anomalies accurately under such scenarios, from the probabilistic perspective, the key question becomes how to learn the normal-data distribution from a contaminated dataset. To this end, we propose to collect two additional small datasets that are comprised of partially-observed normal and anomaly samples, and then use them to help learn the distribution under an adversarial learning scheme. We prove that under some mild conditions, the proposed method is able to learn the correct normal-data distribution. Then, we consider the overfitting issue caused by the small size of the two additional datasets, and a correctness-guaranteed flipping mechanism is further developed to alleviate it. Theoretical results under incomplete observed anomaly types are also presented. Extensive experimental results demonstrate that our method outperforms representative baselines when detecting anomalies under contaminated datasets.

Trailer generation is a challenging video clipping task that aims to select highlighting shots from long videos like movies and re-organize them in an attractive way. In this study, we propose an inverse partial optimal transport (IPOT) framework to achieve music-guided movie trailer generation. In particular, we formulate the trailer generation task as selecting and sorting key movie shots based on audio shots, which involves matching the latent representations across visual and acoustic modalities. We learn a multi-modal latent representation model in the proposed IPOT framework to achieve this aim. In this framework, a two-tower encoder derives the latent representations of movie and music shots, respectively, and an attention-assisted Sinkhorn matching network parameterizes the grounding distance between the shots' latent representations and the distribution of the movie shots. Taking the correspondence between the movie shots and its trailer music shots as the observed optimal transport plan defined on the grounding distances, we learn the model by solving an inverse partial optimal transport problem, leading to a bi-level optimization strategy. We collect real-world movies and their trailers to construct a dataset with abundant label information called CMTD and, accordingly, train and evaluate various automatic trailer generators. Compared with state-of-the-art methods, our IPOT method consistently shows superiority in subjective visual effects and objective quantitative measurements. The code is available at https://github.com/Dixin-Lab/Automatic-Movie-Trailer-Generator.

While following different technical routes, both low-rank and orthogonal adaptation techniques can efficiently adapt large-scale pre-training models in specific tasks or domains based on a small piece of trainable parameters. In this study, we bridge the gap between these two techniques, proposing a simple but effective adaptation method based on Householder reflections. Given a pre-trained model, our method fine-tunes its layers by multiplying each frozen weight matrix with an orthogonal matrix constructed by a chain of learnable Householder reflections (HRs). This HR-based orthogonal fine-tuning is equivalent to an adaptive low-rank adaptation. Moreover, we show that the orthogonality of the reflection planes corresponding to the HRs impacts the model capacity and regularity. The analysis motivates us to regularize the orthogonality of the HRs, leading to different implementations of the proposed Householder reflection adaptation (HRA) method. Compared with state-of-the-art methods, HRA achieves superior performance with fewer learnable parameters when adapting large language models and conditional image generators. The code of the experiments is available at https://github.com/DaShenZi721/HRA, and the method has been merged into the [PEFT](https://github.com/huggingface/peft) package.

Incorporating Search and Recommendation (S&R) services within a singular application is prevalent in online platforms, leading to a new task termed open-app motivation prediction, which aims to predict whether users initiate the application with the specific intent of information searching, or to explore recommended content for entertainment. Studies have shown that predicting users' motivation to open an app can help to improve user engagement and enhance performance in various downstream tasks. However, accurately predicting open-app motivation is not trivial, as it is influenced by user-specific factors, search queries, clicked items, as well as their temporal occurrences. Furthermore, these activities occur sequentially and exhibit intricate temporal dependencies. Inspired by the success of the Neural Hawkes Process (NHP) in modeling temporal dependencies in sequences, this paper proposes a novel neural Hawkes process model to capture the temporal dependencies between historical user browsing and querying actions. The model, referred to as Neural Hawkes Process-based Open-App Motivation prediction model (NHP-OAM), employs a hierarchical transformer and a novel intensity function to encode multiple factors, and open-app motivation prediction layer to integrate time and user-specific information for predicting users' open-app motivations. To demonstrate the superiority of our NHP-OAM model and construct a benchmark for the Open-App Motivation Prediction task, we not only extend the public S&R dataset ZhihuRec but also construct a new real-world Open-App Motivation Dataset (OAMD). Experiments on these two datasets validate NHP-OAM's superiority over baseline models. Further downstream application experiments demonstrate NHP-OAM's effectiveness in predicting users' Open-App Motivation, highlighting the immense application value of NHP-OAM.

When learning graph neural networks (GNNs) in node-level prediction tasks, most existing loss functions are applied for each node independently, even if node embeddings and their labels are non-i.i.d. because of their graph structures. To eliminate such inconsistency, in this study we propose a novel Quasi-Wasserstein (QW) loss with the help of the optimal transport defined on graphs, leading to new learning and prediction paradigms of GNNs. In particular, we design a "Quasi-Wasserstein'' distance between the observed multi-dimensional node labels and their estimations, optimizing the label transport defined on graph edges. The estimations are parameterized by a GNN in which the optimal label transport may determine the graph edge weights optionally. By reformulating the strict constraint of the label transport to a Bregman divergence-based regularizer, we obtain the proposed Quasi-Wasserstein loss associated with two efficient solvers learning the GNN together with optimal label transport. When predicting node labels, our model combines the output of the GNN with the residual component provided by the optimal label transport, leading to a new transductive prediction paradigm. Experiments show that the proposed QW loss applies to various GNNs and helps to improve their performance in node-level classification and regression tasks. The code of this work can be found at https://github.com/SDS-Lab/QW_Loss.

Distribution comparison plays a central role in many machine learning tasks like data classification and generative modeling. In this study, we propose a novel metric, calledHilbert curve projection (HCP) distance, to measure the distance between two probability distributions with low complexity. In particular, we first project two high-dimensional probability distributions using Hilbert curve to obtain a coupling between them, and then calculate the transport distance between these two distributions in the original space, according to the coupling. We show that HCP distance is a proper metric and is well-defined for probability measures with bounded supports. Furthermore, we demonstrate that the modified empirical HCP distance with the$L_{p}$cost in the$d$-dimensional space converges to its population counterpart at a rate of no more than$O(n^{-1/2\max \lbrace d,p\rbrace })$. To suppress the curse-of-dimensionality, we also develop two variants of the HCP distance using (learnable) subspace projections. Experiments on both synthetic and real-world data show that our HCP distance works as an effective surrogate of the Wasserstein distance with low complexity and overcomes the drawbacks of the sliced Wasserstein distance. The code of this work is athttps://github.com/sherlockLitao/HCP.

Providing human-understandable explanations for the matching predictions is still challenging for current legal case matching methods. One difficulty is that legal cases are semi-structured text documents with complicated case-case and case-law article correlations. To tackle the issue, we propose a novel graph optimal transport (GOT)-based legal case matching model that is able to provide not only the matching predictions but also plausible and faithful explanations for the prediction. The model, called GEIOT-Match, first constructs a heterogeneous graph to explicitly represent the semi-structured nature of legal cases and their associations with the law articles. Therefore, matching two legal cases amounts to identifying the rationales from the paired legal case sub-graphs in the heterogeneous graph and then aligning between them. An inverse optimal transport (IOT) model on graphs is learned to extract rationales from paired legal cases. The extracted rationales and the heterogeneous graph demonstrate the key legal characteristics of legal cases, which can be further used to conduct matching and generate explanations for the matching. Experimental results showed that GEIOT-Match outperformed state-of-the-art baselines in terms of matching prediction, rationale extraction, and natural language explanation generation.

As an important sequential model, the temporal point process (TPP) plays a central role in real-world sequence modeling and analysis, whose learning is often based on the maximum likelihood estimation (MLE). However, due to imperfect observations, such as incomplete and sparse sequences that are common in practice, the MLE of TPP models often suffers from overfitting and leads to unsatisfactory generalization power. In this work, we develop a novel hierarchical contrastive (HCL) learning method for temporal point processes, which provides a new regularizer of MLE. In principle, our HCL considers the noise contrastive estimation (NCE) problem at the event-level and at the sequence-level jointly. Given a sequence, the event-level NCE maximizes the probability of each observed event given its history while penalizing the conditional probabilities of the unobserved events. At the same time, we generate positive and negative event sequences from the observed sequence and maximize the discrepancy between their likelihoods through the sequence-level NCE. Instead of using time-consuming simulation methods, we generate the positive and negative sequences via a simple but efficient model-guided thinning process. Experimental results show that the MLE method assisted by the HCL regularizer outperforms classic MLE and other contrastive learning methods in learning various TPP models consistently. The code is available at https://github.com/qingmeiwangdaily/HCL_TPP.

Real-world graphs like social networks are often evolutionary over time, whose observations at different timestamps lead to graph sequences. Modeling such evolutionary graphs is important for many applications, but solving this problem often requires the correspondence between the graphs at different timestamps, which may leak private node information, e.g., the temporal behavior patterns of the nodes. We proposed a Gromov-Wasserstein Autoregressive (GWAR) model to capture the generative mechanisms of evolutionary graphs, which does not require the correspondence information and thus preserves the privacy of the graphs' nodes. This model consists of two autoregressions, predicting the number of nodes and the probabilities of nodes and edges, respectively. The model takes observed graphs as its input and predicts future graphs via solving a joint graph alignment and merging task. This task leads to a fused Gromov-Wasserstein (FGW) barycenter problem, in which we approximate the alignment of the graphs based on a novel inductive fused Gromov-Wasserstein (IFGW) distance. The IFGW distance is parameterized by neural networks and can be learned under mild assumptions, thus, we can infer the FGW barycenters without iterative optimization and predict future graphs efficiently. Experiments show that our GWAR achieves encouraging performance in modeling evolutionary graphs in privacy-preserving scenarios.

Molecular perception aims to construct 3D molecules from 3D atom clouds (i.e., atom types and corresponding 3D coordinates), determining bond connections, bond orders, and other molecular attributes within molecules. It is essential for realizing many applications in cheminformatics and bioinformatics, such as modeling quantum chemistry-derived molecular structures in protein-ligand complexes. Additionally, many molecular generation methods can only generate molecular 3D atom clouds, requiring molecular perception as a necessary post-processing. However, existing molecular perception methods mainly rely on predefined chemical rules and fail to leverage 3D geometric information, whose performance is sub-optimal fully. In this study, we propose MPerformer, an SE(3) Transformer-based molecular perceptron exhibiting SE(3)-invariance, to construct 3D molecules from 3D atom clouds efficiently. Besides, we propose a multi-task pretraining-and-finetuning paradigm to learn this model. In the pretraining phase, we jointly minimize an attribute prediction loss and an atom cloud reconstruction loss, mitigating the data imbalance issue of molecular attributes and enhancing the robustness and generalizability of the model. Experiments show that MPerformer significantly outperforms state-of-the-art molecular perception methods in precision and robustness, benefiting various molecular generation scenarios.

Instead of searching the entire network directly, current NAS approaches increasingly search for multiple relatively small cells to reduce search costs. A major challenge is to jointly measure the similarity of cell micro-architectures and the difference in macro-architectures between different cell-based networks. Recently, optimal transport (OT) has been successfully applied to NAS as it can capture the operational and structural similarity across various networks. However, existing OT-based NAS methods either ignore the cell similarity or focus solely on searching for a single cell architecture. To address these issues, we propose a hierarchical optimal transport metric called HOTNN for measuring the similarity of different networks. In HOTNN, the cell-level similarity computes the OT distance between cells in various networks by considering the similarity of each node and the differences in the information flow costs between node pairs within each cell in terms of operational and structural information. The network-level similarity calculates OT distance between networks by considering both the cell-level similarity and the variation in the global position of each cell within their respective networks. We then explore HOTNN in a Bayesian optimization framework called HOTNAS, and demonstrate its efficacy in diverse tasks. Extensive experiments demonstrate that H OT-NAS can discover network architectures with better performance in multiple modular cell-based search spaces.

Network alignment aims at finding the correspondence of nodes across different networks, which is significant for many applications, e.g., fraud detection and crime network tracing across platforms. In practice, however, accessing the topological information of different networks is often restricted and even forbidden, considering privacy and security issues. Instead, what we observed might be the event sequences of the networks' nodes in the continuous-time domain. In this study, we develop a coupled neural point process-based (CPP) sequence modeling strategy, which provides a solution to privacy-preserving network alignment based on the event sequences. Our CPP consists of a coupled node embedding layer and a neural point process module. The coupled node embedding layer embeds one network's nodes and explicitly models the alignment matrix between the two networks. Accordingly, it parameterizes the node embeddings of the other network by the push-forward operation. Given the node embeddings, the neural point process module jointly captures the dynamics of the two networks' event sequences. We learn the CPP model in a maximum likelihood estimation framework with an inverse optimal transport (IOT) regularizer. Experiments show that our CPP is compatible with various point process backbones and is robust to the model misspecification issue, which achieves encouraging performance on network alignment. The code is available at https://github.com/Dixin-s-Lab/CNPP.

As a fundamental problem in Operations Research, sparse process flexibility design (SPFD) aims to design a manufacturing network across industries that achieves a trade-off between the efficiency and robustness of supply chains. In this study, we propose a novel solution to this problem with the help of computational optimal transport techniques. Given a set of supply-demand pairs, we formulate the SPFD task approximately as a group sparse optimal transport (GSOT) problem, in which a group of couplings between the supplies and demands is optimized with a group sparse regularizer. We solve this optimization problem via an algorithmic framework of alternating direction method of multipliers (ADMM), in which the target network topology is updated by soft-thresholding shrinkage, and the couplings of the OT problems are updated via a smooth OT algorithm in parallel. This optimization algorithm has guaranteed convergence and provides a generalized framework for the SPFD task, which is applicable regardless of whether the supplies and demands are balanced. Experiments show that our GSOT-based method can outperform representative heuristic methods in various SPFD tasks. Additionally, when implementing the GSOT method, the proposed ADMM-based optimization algorithm is comparable or superior to the commercial software Gurobi. The code is available at https://github.com/Dixin-s-Lab/GSOT.

Video summarization is a critical task in video analysis that aims to create a brief yet informative summary of the original video (i.e., a set of keyframes) while retaining its primary content. Supervised summarization methods rely on time-consuming keyframe labeling and thus often suffer from the insufficiency issue of training data. In contrast, the performance of unsupervised summarization methods is often unsatisfactory due to the lack of semantically-meaningful guidance on the keyframe selection. In this study, we propose a novel self-supervised video summarization framework with the help of computational optimal transport techniques. Specifically, we generate textual descriptions from video shots and learn the projection from the textual embeddings to the visual ones together with an optimal transport plan between them via solving an inverse optimal transport problem. We propose an alternating optimization algorithm to solve this problem efficiently and design an effective mechanism in the algorithm to avoid trivial solutions. Given the optimal transport plan and the underlying distance between the projected textual embeddings and the visual ones, we synthesize pseudo-significance scores for video frames and leverage the scores as offline supervision to train a keyframe selector. Without subjective and error-prone manual annotations, the proposed framework surpasses previous unsupervised methods in producing high-quality results for generic and instructional video summarization tasks, whose performance even is comparable to those supervised competitors. The code is available at https://github.com/Dixin-s-Lab/Video-Summary-IOT.

Temporal event sequences associated with different event types (e.g., location indices, disease types) are observed in various applications such as disaster resilience, criminology, and healthcare. Temporal point processes (TPPs) have been developed to capture the exciting patterns between events and forecast future events quantitatively. Unfortunately, the events with different types often suffer from unknown biased observations in real-world scenarios due to external interference. Accordingly, the temporal point processes learned by conventional maximum likelihood estimation (MLE) from such biased data may be misspecified and may lead to inaccurate predictions. To overcome this issue, we model biased event sequences as modulating TPPs with additional unknown thinning processes. Furthermore, we develop a novel debiased imitation learning framework to learn the modulated TPPs and suppress the negative influences of biased data, which is more robust than conventional MLE. When applying the debiased imitation learning framework, we design a simple but effective reward function based on the historical embedding obtained by the TPP model. Experiments on three real-world datasets demonstrate that our proposed method significantly outperforms existing methods.

Traditional multi-view learning methods often rely on two assumptions: ( i) the samples in different views are well-aligned, and ( ii) their representations obey the same distribution in a latent space. Unfortunately, these two assumptions may be questionable in practice, which limits the application of multi-view learning. In this work, we propose a differentiable hierarchical optimal transport (DHOT) method to mitigate the dependency of multi-view learning on these two assumptions. Given arbitrary two views of unaligned multi-view data, the DHOT method calculates the sliced Wasserstein distance between their latent distributions. Based on these sliced Wasserstein distances, the DHOT method further calculates the entropic optimal transport across different views and explicitly indicates the clustering structure of the views. Accordingly, the entropic optimal transport, together with the underlying sliced Wasserstein distances, leads to a hierarchical optimal transport distance defined for unaligned multi-view data, which works as the objective function of multi-view learning and leads to a bi-level optimization task. Moreover, our DHOT method treats the entropic optimal transport as a differentiable operator of model parameters. It considers the gradient of the entropic optimal transport in the backpropagation step and thus helps improve the descent direction for the model in the training phase. We demonstrate the superiority of our bi-level optimization strategy by comparing it to the traditional alternating optimization strategy. The DHOT method is applicable for both unsupervised and semi-supervised learning. Experimental results show that our DHOT method is at least comparable to state-of-the-art multi-view learning methods on both synthetic and real-world tasks, especially for challenging scenarios with unaligned multi-view data.

Real-world 3D structured data like point clouds and skeletons often can be represented as data in a 3D rotation group (denoted as [Formula: see text]). However, most existing neural networks are tailored for the data in the euclidean space, which makes the 3D rotation data not closed under their algebraic operations and leads to sub-optimal performance in 3D-related learning tasks. To resolve the issues caused by the above mismatching between data and model, we propose a novel non-real neuron model called quaternion product unit (QPU) to represent data on 3D rotation groups. The proposed QPU leverages quaternion algebra and the law of the 3D rotation group, representing 3D rotation data as quaternions and merging them via a weighted chain of Hamilton products. We demonstrate that the QPU mathematically maintains the [Formula: see text] structure of the 3D rotation data during the inference process and disentangles the 3D representations into "rotation-invariant" features and "rotation-equivariant" features, respectively. Moreover, we design a fast QPU to accelerate the computation of QPU. The fast QPU applies a tree-structured data indexing process, and accordingly, leverages the power of parallel computing, which reduces the computational complexity of QPU in a single thread from O(N) to O(logN). Taking the fast QPU as a basic module, we develop a series of quaternion neural networks (QNNs), including quaternion multi-layer perceptron (QMLP), quaternion message passing (QMP), and so on. In addition, we make the QNNs compatible with conventional real-valued neural networks and applicable for both skeletons and point clouds. Experiments on synthetic and real-world 3D tasks show that the QNNs based on our fast QPUs are superior to state-of-the-art real-valued models, especially in the scenarios requiring the robustness to random rotations. The code of this work is available at https://github.com/SuferQin/Fast-QPU.

Global pooling is one of the most significant operations in many machine learning models and tasks, which works for information fusion and structured data (like sets and graphs) representation. However, without solid mathematical fundamentals, its practical implementations often depend on empirical mechanisms and thus lead to sub-optimal, even unsatisfactory performance. In this work, we develop a novel and generalized global pooling framework through the lens of optimal transport. The proposed framework is interpretable from the perspective of expectation-maximization. Essentially, it aims at learning an optimal transport across sample indices and feature dimensions, making the corresponding pooling operation maximize the conditional expectation of input data. We demonstrate that most existing pooling methods are equivalent to solving a regularized optimal transport (ROT) problem with different specializations, and more sophisticated pooling operations can be implemented by hierarchically solving multiple ROT problems. Making the parameters of the ROT problem learnable, we develop a family of regularized optimal transport pooling (ROTP) layers. We implement the ROTP layers as a new kind of deep implicit layer. Their model architectures correspond to different optimization algorithms. We test our ROTP layers in several representative set-level machine learning scenarios, including multi-instance learning (MIL), graph classification, graph set representation, and image classification. Experimental results show that applying our ROTP layers can reduce the difficulty of the design and selection of global pooling - our ROTP layers may either imitate some existing global pooling methods or lead to some new pooling layers fitting data better.

Graph representation is a challenging and significant problem for many real-world applications. In this work, we propose a novel paradigm called "Gromov-Wasserstein Factorization" (GWF) to learn graph representations in a flexible and interpretable way. Given a set of graphs, whose correspondence between nodes is unknown and whose sizes can be different, our GWF model reconstructs each graph by a weighted combination of some "graph factors" under a pseudo-metric called Gromov-Wasserstein (GW) discrepancy. This model leads to a new nonlinear factorization mechanism of the graphs. The graph factors are shared by all the graphs, which represent the typical patterns of the graphs' structures. The weights associated with each graph indicate the graph factors' contributions to the graph's reconstruction, which lead to a permutation-invariant graph representation. We learn the graph factors of the GWF model and the weights of the graphs jointly by minimizing the overall reconstruction error. When learning the model, we reparametrize the graph factors and the weights to unconstrained model parameters and simplify the backpropagation of gradient with the help of the envelope theorem. For the GW discrepancy (the critical training step), we consider two algorithms to compute it, which correspond to the proximal point algorithm (PPA) and Bregman alternating direction method of multipliers (BADMM), respectively. Furthermore, we propose some extensions of the GWF model, including (i) combining with a graph neural network and learning graph representations in an auto-encoding manner, (ii) representing the graphs with node attributes, and (iii) working as a regularizer for semi-supervised graph classification. Experiments on various datasets demonstrate that our GWF model is comparable to the state-of-the-art methods. The graph representations derived by it perform well in graph clustering and classification tasks.

Sequential recommendation has recently attracted increasing attention from the industry and academic communities. While previous models have achieved remarkable successes, an important problem may still hinder their performances, that is, the sparsity of the real-world data. In this paper, we propose a novel counterfactual data augmentation framework to alleviate the problem of data sparsity. In specific, our framework contains a sampler model and an anchor model. The sampler model aims to generate high-quality user behavior sequences, while the anchor model is trained based on the original and new generated samples, and leveraged to provide the final recommendation list. To implement the sampler model, we first design four types of heuristic methods based on either random or frequency-based strategies. And then, to improve the quality of the generated sequences, we propose two learning-based samplers by discovering the decision boundaries or increasing the sample informativeness. At last, we build an RL based model to automatically determine where to edit the history behaviors and how many items should be replaced. Considering that the sampler model can be imperfect, we, at last, analyze the influence of the noisy information contained in the generated sequences on the anchor model in theory, and design a simple but effective method to better serve the anchor model. We conduct extensive experiments to demonstrate the effectiveness of our model.

Temporal data are ubiquitous in real-world applications, and they can be generally divided into two categories: 1) synchronous temporal data which are basically equivalent to time series data; and 2) the asynchronous data which are often in the form of event data with a time stamp in continuous time-space. In fact, the event data are often converted to the time series by aggregating the event count in equal time intervals in many previous approaches. While it is often of one’s greater interest to directly establish models based on the raw event data whose time stamps carry useful information, especially for those time-sensitive tasks, ranging from earthquake prediction, crime analysis, to infectious disease diffusion forecasting, etc. Developing the spatio-temporal point process and the related applications is the theme of this Special Issue, which treats an event as a point in the spatio-temporal space, with possibly extra attributes. The model captures the instantaneous happening rate of the events and their potential dependency. The derived use cases often refer to future events prediction, and causality estimation.

When applying multi-instance learning (MIL) to make predictions for bags of instances, the prediction accuracy of an instance often depends on not only the instance itself but also its context in the corresponding bag. From the viewpoint of causal inference, such bag contextual prior works as a confounder and may result in model robustness and interpretability issues. Focusing on this problem, we propose a novel interventional multi-instance learning (IMIL) framework to achieve deconfounded instance-level prediction. Unlike traditional likelihood-based strategies, we design an Expectation-Maximization (EM) algorithm based on causal intervention, providing a robust instance selection in the training phase and suppressing the bias caused by the bag contextual prior. Experiments on pathological image analysis demonstrate that our IMIL method substantially reduces false positives and outperforms state-of-the-art MIL methods.

Masked autoencoder (MAE), as an effective self-supervised learner for computer vision and natural language processing, has been recently applied to molecule representation learning. In this paper, we identify two issues in applying MAE to pre-train Transformer-based models on molecular graphs that existing works have ignored. (1) As only atoms are abstracted as tokens and then reconstructed, the chemical bonds are not decided in the decoded molecule, making molecules with different arrangements of the same atoms indistinguishable. (2) Although a high mask ratio that corresponds to a challenging reconstruction task has been proved beneficial in the vision domain, it cannot be trivially leveraged on molecular graphs as there is less redundancy of information in graph data. To resolve these issues, we propose a novel framework, Molecular Graph Mask AutoEncoder (MGMAE). As the first step in MGMAE, we transform each molecular graph into a heterogeneous atom-bond graph to fully use the bond attributes and design unidirectional position encoding for such graphs. Then we propose a hybrid masking mechanism that exploits the complementary nature between atoms' attributive and spatial features. Meanwhile, we compensate for the mask embedding by a dynamic aggregation representation that exploits the correlations between topologically adjacent tokens. As a result, MGMAE can reconstruct the masked atoms, the masked bonds, and the relative distance among atoms simultaneously, with a high mask ratio. We compare MGMAE with the state-of-the-art methods on various molecular benchmarks and show the competitiveness of MGMAE in both regression and classification tasks.

Multi-modal clustering aims at finding a clustering structure shared by the data of different modalities in an unsupervised way. Currently, solving this problem often relies on two assumptions: i) the multi-modal data own the same latent distribution, and ii) the observed multi-modal data are well-aligned and without any missing modalities. Unfortunately, these two assumptions are often questionable in practice and thus limit the feasibility of many multi-modal clustering methods. In this work, we develop a new multi-modal clustering method based on the Gromovization of optimal transport distance, which relaxes the dependence on the above two assumptions. In particular, given the data of different modalities, whose correspondence is unknown, our method learns the Gromov-Wasserstein (GW) barycenter of their kernel matrices. Driven by the modularity maximization principle, the GW barycenter helps to explore the clustering structure shared by different modalities. Moreover, the GW barycenter is associated with the GW distances between the different modalities to the clusters, and the optimal transport plans corresponding to the GW distances help to achieve the alignment and the clustering of the multi-modal data jointly. Experimental results show that our method outperforms state-of-the-art multi-modal clustering methods, especially when the data are (partially or completely) unaligned. The code is available at https://github.com/rucnyz/GWMAC.

Poster generation is a significant task for a wide range of applications, which is often time-consuming and requires lots of manual editing and artistic experience. In this paper, we propose a novel data-driven framework, called Text2Poster, to automatically generate visually-effective posters from textual information. Imitating the process of manual poster editing, our framework leverages a large-scale pretrained visual-textual model to retrieve background images from given texts, lays out the texts on the images iteratively by cascaded autoencoders, and finally, stylizes the texts by a matching-based method. We learn the modules of the framework by weakly-and self-supervised learning strategies, mitigating the demand for labeled data. Both objective and subjective experiments demonstrate that our Text2Poster outperforms state-of-the-art methods, including academic research and commercial software, on the quality of generated posters.

The prediction of protein-ligand binding affinity is of great significance for discovering lead compounds in drug research. Facing this challenging task, most existing prediction methods rely on the topological and/or spatial structure of molecules and the local interactions while ignoring the multi-level inter-molecular interactions between proteins and ligands, which often lead to sub-optimal performance. To solve this issue, we propose a novel global-local interaction (GLI) framework to predict protein-ligand binding affinity. In particular, our GLI framework considers the inter-molecular interactions between proteins and ligands, which involve not only the high-energy short-range interactions between closed atoms but also the low-energy long-range interactions between non-bonded atoms. For each pair of protein and ligand, our GLI embeds the long-range interactions globally and aggregates local short-range interactions, respectively. Such a joint global-local interaction modeling strategy helps to improve prediction accuracy, and the whole framework is compatible with various neural network-based modules. Experiments demonstrate that our GLI framework outperforms state-of-the-art methods with simple neural network architectures and moderate computational costs.

Temporal action alignment aims at segmenting videos into clips and tagging each clip with a textual description, which is an important task of video semantic analysis. Most existing methods, however, rely on supervised learning to train their alignment models, whose applications are limited because of the common insufficiency issue of labeled videos. To mitigate this issue, we propose a weakly-supervised temporal action alignment method based on a novel computational optimal transport technique called unbalanced spectral fused Gromov-Wasserstein (US-FGW) distance. Instead of using videos with known clips and corresponding textual tags, our method just needs each training video to be associated with a set of (unsorted) texts while does not require the fine-grained correspondence between the frames and the texts. Given such weakly-supervised video-text pairs, our method trains the representation models of the video frames and the texts jointly in a probabilistic or deterministic autoencoding architecture and penalizes the US-FGW distance between the distribution of visual latent codes and that of textual latent codes. We compute the US-FGW distance efficiently by leveraging the Bregman ADMM algorithm. Furthermore, we generalize classic contrastive learning framework and reformulate it based on the proposed US-FGW distance, which provides a new viewpoint of contrastive learning for our problem. Experimental results show that our method and its variants outperform state-of-the-art weakly-supervised temporal action alignment methods, whose results are even comparable to those derived by supervised learning methods on some specific evaluation measurements. The code is available at \urlhttps://github.com/hhhh1138/Temporal-Action-Alignment-USFGW.

As an essential operation of legal retrieval, legal case matching plays a central role in intelligent legal systems. This task has a high demand on the explainability of matching results because of its critical impacts on downstream applications --- the matched legal cases may provide supportive evidence for the judgments of target cases and thus influence the fairness and justice of legal decisions. Focusing on this challenging task, we propose a novel and explainable method, namely IOT-Match, with the help of computational optimal transport, which formulates the legal case matching problem as an inverse optimal transport (IOT) problem. Different from most existing methods, which merely focus on the sentence-level semantic similarity between legal cases, our IOT-Match learns to extract rationales from paired legal cases based on both semantics and legal characteristics of their sentences. The extracted rationales are further applied to generate faithful explanations and conduct matching. Moreover, the proposed IOT-Match is robust to the alignment label insufficiency issue commonly in practical legal case matching tasks, which is suitable for both supervised and semi-supervised learning paradigms. To demonstrate the superiority of our IOT-Match method and construct a benchmark of explainable legal case matching task, we not only extend the well-known Challenge of AI in Law (CAIL) dataset but also build a new Explainable Legal cAse Matching (ELAM) dataset, which contains lots of legal cases with detailed and explainable annotations. Experiments on these two datasets show that our IOT-Match outperforms state-of-the-art methods consistently on matching prediction, rationale extraction, and explanation generation.

We propose a novel and principled method to learn a nonparametric graph model called graphon, which is defined in an infinite-dimensional space and represents arbitrary-size graphs. Based on the weak regularity lemma from the theory of graphons, we leverage a step function to approximate a graphon. We show that the cut distance of graphons can be relaxed to the Gromov-Wasserstein distance of their step functions. Accordingly, given a set of graphs generated by an underlying graphon, we learn the corresponding step function as the Gromov-Wasserstein barycenter of the given graphs. Furthermore, we develop several enhancements and extensions of the basic algorithm, e.g., the smoothed Gromov-Wasserstein barycenter for guaranteeing the continuity of the learned graphons and the mixed Gromov-Wasserstein barycenters for learning multiple structured graphons. The proposed approach overcomes drawbacks of prior state-of-the-art methods, and outperforms them on both synthetic and real-world data. The code is available at https://github.com/HongtengXu/SGWB-Graphon.

Many representative graph neural networks, $e.g.$, GPR-GNN and ChebNet, approximate graph convolutions with graph spectral filters. However, existing work either applies predefined filter weights or learns them without necessary constraints, which may lead to oversimplified or ill-posed filters. To overcome these issues, we propose $\textit{BernNet}$, a novel graph neural network with theoretical support that provides a simple but effective scheme for designing and learning arbitrary graph spectral filters. In particular, for any filter over the normalized Laplacian spectrum of a graph, our BernNet estimates it by an order-$K$ Bernstein polynomial approximation and designs its spectral property by setting the coefficients of the Bernstein basis. Moreover, we can learn the coefficients (and the corresponding filter weights) based on observed graphs and their associated signals and thus achieve the BernNet specialized for the data. Our experiments demonstrate that BernNet can learn arbitrary spectral filters, including complicated band-rejection and comb filters, and it achieves superior performance in real-world graph modeling tasks. Code is available at https://github.com/ivam-he/BernNet.

Sequential recommendation aims at predicting users' preferences based on their historical behaviors. However, this recommendation strategy may not perform well in practice due to the sparsity of the real-world data. In this paper, we propose a novel counterfactual data augmentation framework to mitigate the impact of the imperfect training data and empower sequential recommendation models. Our framework is composed of a sampler model and an anchor model. The sampler model aims to generate new user behavior sequences based on the observed ones, while the anchor model is leveraged to provide the final recommendation list, which is trained based on both observed and generated sequences. We design the sampler model to answer the key counterfactual question: "what would a user like to buy if her previously purchased items had been different?". Beyond heuristic intervention methods, we leverage two learning-based methods to implement the sampler model, and thus, improve the quality of the generated sequences when training the anchor model. Additionally, we analyze the influence of the generated sequences on the anchor model in theory and achieve a trade-off between the information and the noise introduced by the generated sequences. Experiments on nine real-world datasets demonstrate our framework's effectiveness and generality.

We propose an optimal transport (OT) framework for generalized zero-shot learning (GZSL), seeking to distinguish samples for both seen and unseen classes, with the assist of auxiliary attributes. The discrepancy between features and attributes is minimized by solving an optimal transport problem. Specifically, we build a conditional generative model to generate features from seen-class attributes, and establish an optimal transport between the distribution of the generated features and that of the real features. The generative model and the optimal transport are optimized iteratively with an attribute-based regularizer, that further enhances the discriminative power of the generated features. A classifier is learned based on the features generated for both the seen and unseen classes. In addition to generalized zero-shot learning, our framework is also applicable to standard and transductive ZSL problems. Experiments show that our optimal transport-based method outperforms state-of-the-art methods on several benchmark datasets.

In many situations, such as transferring data between devices and recording precious moments, we would like to capture the contents on screens using digital cameras for convenience. These recaptured screen images and videos suffer from a special type of degradation called “moiré pattern”, which is caused by the aliasing between the grid of display screen and the array of camera sensor. However, few works are proposed to tackle this problem. Considering the great success of convolutional neural networks (CNNs) in image restoration, we propose a CNN-based moiré removal method for recaptured screen images. There are mainly two contributions in this paper. First, for the generation of training data, we propose an image registration algorithm via global homography transform and local patch matching to compensate the significant viewpoint disparity between the recaptured screen image and the moiré-free image obtained via screenshot. We construct a moiré removal and brightness improvement (MRBI) database with aligned moiré-free and moiré images. Second, we propose a convolutional neural Network with Additive and Multiplicative modules (termed as AMNet) to transfer the low light moiré image to the bright moiré-free image. The proposed network is trained with pixel-wise loss, perceptual loss, and adversarial loss. Extensive experiments on 340 test images demonstrate that the proposed method outperforms state-of-the-art moiré removal methods.

We propose a novel graph-driven generative model, that unifies multiple heterogeneous learning tasks into the same framework. The proposed model is based on the fact that heterogeneous learning tasks, which correspond to different generative processes, often rely on data with a shared graph structure. Accordingly, our model combines a graph convolutional network (GCN) with multiple variational autoencoders, thus embedding the nodes of the graph (i.e., samples for the tasks) in a uniform manner, while specializing their organization and usage to different tasks. With a focus on healthcare applications (tasks), including clinical topic modeling, procedure recommendation and admission-type prediction, we demonstrate that our method successfully leverages information across different tasks, boosting performance in all tasks and outperforming existing state-of-the-art approaches.

We propose a new nonlinear factorization model for graphs that are with topological structures, and optionally, node attributes. This model is based on a pseudometric called Gromov-Wasserstein (GW) discrepancy, which compares graphs in a relational way. It estimates observed graphs as GW barycenters constructed by a set of atoms with different weights. By minimizing the GW discrepancy between each observed graph and its GW barycenter-based estimation, we learn the atoms and their weights associated with the observed graphs. The model achieves a novel and flexible factorization mechanism under GW discrepancy, in which both the observed graphs and the learnable atoms can be unaligned and with different sizes. We design an effective approximate algorithm for learning this Gromov-Wasserstein factorization (GWF) model, unrolling loopy computations as stacked modules and computing gradients with backpropagation. The stacked modules can be with two different architectures, which correspond to the proximal point algorithm (PPA) and Bregman alternating direction method of multipliers (BADMM), respectively. Experiments show that our model obtains encouraging results on clustering graphs.

We propose a novel quaternion product unit (QPU) to represent data on 3D rotation groups. The QPU leverages quaternion algebra and the law of 3D rotation group, representing 3D rotation data as quaternions and merging them via a weighted chain of Hamilton products. We prove that the representations derived by the proposed QPU can be disentangled into "rotation-invariant" features and "rotation-equivariant" features, respectively, which supports the rationality and the efficiency of the QPU in theory. We design quaternion neural networks based on our QPUs and make our models compatible with existing deep learning models. Experiments on both synthetic and real-world data show that the proposed QPU is beneficial for the learning tasks requiring rotation robustness.

We propose a new algorithmic framework for learning autoencoders of data distributions. In this framework, we minimize the discrepancy between the model distribution and the target one, with relational regularization on learnable latent prior. This regularization penalizes the fused Gromov-Wasserstein (FGW) distance between the latent prior and its corresponding posterior, which allows us to learn a structured prior distribution associated with the generative model in a flexible way. Moreover, it helps us co-train multiple autoencoders even if they are with heterogeneous architectures and incomparable latent spaces. We implement the framework with two scalable algorithms, making it applicable for both probabilistic and deterministic autoencoders. Our relational regularized autoencoder (RAE) outperforms existing methods, e.g., variational autoencoder, Wasserstein autoencoder, and their variants, on generating images. Additionally, our relational co-training strategy of autoencoders achieves encouraging results in both synthesis and real-world multi-view learning tasks.

A novel single-image rain removal method is proposed based on multi-scale cascading image generation (MSCG). In particular, the proposed method consists of an encoder extracting multi-scale features from images and a decoder generating de-rained images with a cascading mechanism. The encoder ensembles the convolution neural networks using the kernels with different sizes, and integrates their outputs across different scales. The decoder implements a coarse-to-fine image generation framework, adding fine details incrementally to the final de-rained images according to the spatial contextual information on different scales. We test the proposed method on both synthetic and real-world datasets. Experimental results show that the proposed method is robust to the changes of scene, e.g., the viewpoint and the depth, the heaviness of rain, etc., which suppresses the blurring problem of de-rained image and outperforms state-of-the-art methods consistently.

A novel Gromov-Wasserstein learning framework is proposed to jointly match (align) graphs and learn embedding vectors for the associated graph nodes. Using Gromov-Wasserstein discrepancy, we measure the dissimilarity between two graphs and find their correspondence, according to the learned optimal transport. The node embeddings associated with the two graphs are learned under the guidance of the optimal transport, the distance of which not only reflects the topological structure of each graph but also yields the correspondence across the graphs. These two learning steps are mutually-beneficial, and are unified here by minimizing the Gromov-Wasserstein discrepancy with structural regularizers. This framework leads to an optimization problem that is solved by a proximal point method. We apply the proposed method to matching problems in real-world networks, and demonstrate its superior performance compared to alternative approaches.

Real-world entities' behaviors, associated with their side information, are often recorded over time as asynchronous event sequences. Such event sequences are the basis of many practical applications, neural spiking train study, earth quack prediction, crime analysis, infectious disease diffusion forecasting, condition-based preventative maintenance, information retrieval and behavior-based network analysis and services, etc. Temporal point process (TPP) is a principled mathematical tool for the modeling and learning of asynchronous event sequences, which captures the instantaneous happening rate of the events and the temporal dependency between historical and current events. TPP provides us with an interpretable model to describe the generative mechanism of event sequences, which is beneficial for event prediction and causality analysis. Recently, it has been shown that TPP has potentials to many machine learning and data science applications and can be combined with other cutting-edge machine learning techniques like deep learning, reinforcement learning, adversarial learning, and so on.

We propose a scalable Gromov-Wasserstein learning (S-GWL) method and establish a novel and theoretically-supported paradigm for large-scale graph analysis. The proposed method is based on the fact that Gromov-Wasserstein discrepancy is a pseudometric on graphs. Given two graphs, the optimal transport associated with their Gromov-Wasserstein discrepancy provides the correspondence between their nodes and achieves graph matching. When one of the graphs is a predefined graph with isolated but self-connected nodes ($i.e.$, disconnected graph), the optimal transport indicates the clustering structure of the other graph and achieves graph partitioning. Further, we extend our method to multi-graph partitioning and matching by learning a Gromov-Wasserstein barycenter graph for multiple observed graphs. Our method combines a recursive $K$-partition mechanism with a warm-start proximal gradient algorithm, whose time complexity is $\mathcal{O}(K(E+V)\log_K V)$ for graphs with $V$ nodes and $E$ edges. To our knowledge, our method is the first attempt to make Gromov-Wasserstein discrepancy applicable to large-scale graph analysis and unify graph partitioning and matching into the same framework. It outperforms state-of-the-art graph partitioning and matching methods, achieving a trade-off between accuracy and efficiency.

Fashion recommendation has attracted increasing attention from both industry and academic communities. This paper proposes a novel neural architecture for fashion recommendation based on both image region-level features and user review information. Our basic intuition is that: for a fashion image, not all the regions are equally important for the users, i.e., people usually care about a few parts of the fashion image. To model such human sense, we learn an attention model over many pre-segmented image regions, based on which we can understand where a user is really interested in on the image, and correspondingly, represent the image in a more accurate manner. In addition, by discovering such fine-grained visual preference, we can visually explain a recommendation by highlighting some regions of its image. For better learning the attention model, we also introduce user review information as a weak supervision signal to collect more comprehensive user preference. In our final framework, the visual and textual features are seamlessly coupled by a multimodal attention network. Based on this architecture, we can not only provide accurate recommendation, but also can accompany each recommended item with novel visual explanations. We conduct extensive experiments to demonstrate the superiority of our proposed model in terms of Top-N recommendation, and also we build a collectively labeled dataset for evaluating our provided visual explanations in a quantitative manner.

Integrating external knowledge into the recommendation system has attracted increasing attention in both industry and academic communities. Recent methods mostly take the power of neural network for effective knowledge representation to improve the recommendation performance. However, the heavy deep architectures in existing models are usually incorporated in an embedded manner, which may greatly increase the model complexity and lower the runtime efficiency. To simultaneously take the power of deep learning for external knowledge modeling as well as maintaining the model efficiency at test time, we reformulate the problem of recommendation with external knowledge into a generalized distillation framework . The general idea is to free the complex deep architecture into a separate model, which is only used in the training phrase, while abandoned at test time. In particular, in the training phrase, the external knowledge is processed by a comprehensive teacher model to produce valuable information to teach a simple and efficient student model. Once the framework is learned, the teacher model is abandoned, and only the succinct yet enhanced student model is used to make fast predictions at test time. In this article, we specify the external knowledge as user review, and to leverage it in an effective manner, we further extend the traditional generalized distillation framework by designing a Selective Distillation Network (SDNet) with adversarial adaption and orthogonality constraint strategies to make it more robust to noise information. Extensive experiments verify that our model can not only improve the performance of rating prediction, but also can significantly reduce time consumption when making predictions as compared with several state-of-the-art methods.

Estimating the future event sequence conditioned on current observations is a long-standing and challenging task in temporal analysis. On one hand for many real-world problems the underlying dynamics can be very complex and often unknown. This renders the traditional parametric point process models often fail to fit the data for their limited capacity. On the other hand, long-term prediction suffers from the problem of bias exposure where the error accumulates and propagates to future prediction. Our new model builds upon the sequence to sequence (seq2seq) prediction network. Compared with parametric point process models, its modeling capacity is higher and has better flexibility for fitting real-world data. The main novelty of the paper is to mitigate the second challenge by introducing the likelihood-free loss based on Wasserstein distance between point processes, besides negative maximum likelihood loss used in the traditional seq2seq model. Wasserstein distance, unlike KL divergence i.e. MLE loss, is sensitive to the underlying geometry between samples and can robustly enforce close geometry structure between them. This technique is proven able to improve the vanilla seq2seq model by a notable margin on various tasks.

The superposition of temporal point processes has been studied for many years, although the usefulness of such models for practical applications has not be fully developed. We investigate superposed Hawkes process as an important class of such models, with properties studied in the framework of least squares estimation. The superposition of Hawkes processes is demonstrated to be beneficial for tightening the upper bound of excess risk under certain conditions, and we show the feasibility of the benefit in typical situations. The usefulness of superposed Hawkes processes is verified on synthetic data, and its potential to solve the cold-start problem of recommendation systems is demonstrated on real-world data.

Transferring a low-dynamic-range (LDR) image to a high-dynamic-range (HDR) image, which is the so-called inverse tone mapping (iTM), is an important imaging technique to improve visual effects of imaging devices. In this paper, we propose a novel deep learning-based iTM method, which learns an inverse tone mapping network with a generative adversarial regularizer. In the framework of alternating optimization, we learn a U-Net-based HDR image generator to transfer input LDR images to HDR ones, and a simple CNN-based discriminator to classify the real HDR images and the generated ones. Specifically, when learning the generator we consider the content-related loss and the generative adversarial regularizer jointly to improve the stability and the robustness of the generated HDR images. Using the learned generator as the proposed inverse tone mapping network, we achieve superior iTM results to the state-of-the-art methods consistently.

How to effectively approximate real-valued parameters with binary codes plays a central role in neural network binarization. In this work, we reveal an important fact that binarizing different layers has a widely varied effect on the compression ratio of network and the loss of performance. Based on this fact, we propose a novel and flexible neural network binarization method by introducing the concept of layer-wise priority which binarizes parameters in inverse order of their layer depth. In each training step, our method selects a specific network layer, minimizes the discrepancy between the original real-valued weights and its binary approximations following block coordinate descent scheme. During the iteration of the above process, it is significant that we can flexibly decide whether to binarize the remaining floating layers or not and explore a trade-off between the loss of performance and the compression ratio of model. The resulting binary network is applied for efficient pedestrian detection. Experimental results on several benchmarks show that under the same compression ratio, the model compressed by our method achieves much lower miss rate and faster detection speed than those obtained by the state-of-the-art neural network binarization method.

A parametric point process model is developed, with modeling based on the assumption that sequential observations often share latent phenomena, while also possessing idiosyncratic effects. An alternating optimization method is proposed to learn a “registered” point process that accounts for shared structure, as well as “warping” functions that characterize idiosyncratic aspects of each observed sequence. Under reasonable constraints, in each iteration we update the sample-specific warping functions by solving a set of constrained nonlinear programming problems in parallel, and update the model by maximum likelihood estimation. The justifiability, complexity and robustness of the proposed method are investigated in detail, and the influence of sequence stitching on the learning results is examined empirically. Experiments on both synthetic and real-world data demonstrate that the method yields explainable point process models, achieving encouraging results compared to state-of-the-art methods.

A continuous-time tensor factorization method is developed for event sequences containing multiple "modalities." Each data element is a point in a tensor, whose dimensions are associated with the discrete alphabet of the modalities. Each tensor data element has an associated time of occurence and a feature vector. We model such data based on pairwise interactive point processes, and the proposed framework connects pairwise tensor factorization with a feature-embedded point process. The model accounts for interactions within each modality, interactions across different modalities, and continuous-time dynamics of the interactions. Model learning is formulated as a convex optimization problem, based on online alternating direction method of multipliers. Compared to existing state-of-the-art methods, our approach captures the latent structure of the tensor and its evolution over time, obtaining superior results on real-world datasets.

We propose a novel Wasserstein method with a distillation mechanism, yielding joint learning of word embeddings and topics. The proposed method is based on the fact that the Euclidean distance between word embeddings may be employed as the underlying distance in the Wasserstein topic model. The word distributions of topics, their optimal transport to the word distributions of documents, and the embeddings of words are learned in a unified framework. When learning the topic model, we leverage a distilled ground-distance matrix to update the topic distributions and smoothly calculate the corresponding optimal transports. Such a strategy provides the updating of word embeddings with robust guidance, improving algorithm convergence. As an application, we focus on patient admission records, in which the proposed method embeds the codes of diseases and procedures and learns the topics of admissions, obtaining superior performance on clinically-meaningful disease network construction, mortality prediction as a function of admission codes, and procedure recommendation.

User preferences are usually dynamic in real-world recommender systems, and a user»s historical behavior records may not be equally important when predicting his/her future interests. Existing recommendation algorithms -- including both shallow and deep approaches -- usually embed a user»s historical records into a single latent vector/representation, which may have lost the per item- or feature-level correlations between a user»s historical records and future interests. In this paper, we aim to express, store, and manipulate users» historical records in a more explicit, dynamic, and effective manner. To do so, we introduce the memory mechanism to recommender systems. Specifically, we design a memory-augmented neural network (MANN) integrated with the insights of collaborative filtering for recommendation. By leveraging the external memory matrix in MANN, we store and update users» historical records explicitly, which enhances the expressiveness of the model. We further adapt our framework to both item- and feature-level versions, and design the corresponding memory reading/writing operations according to the nature of personalized recommendation scenarios. Compared with state-of-the-art methods that consider users» sequential behavior for recommendation, e.g., sequential recommenders with recurrent neural networks (RNN) or Markov chains, our method achieves significantly and consistently better performance on four real-world datasets. Moreover, experimental analyses show that our method is able to extract the intuitive patterns of how users» future actions are affected by previous behaviors.

Fractal analysis has been widely used in computer vision, especially in texture image processing and texture analysis. The key concept of fractal-based image model is the fractal dimension, which is invariant to bi-Lipschitz transformation of image, and thus capable of representing intrinsic structural information of image robustly. However, the invariance of fractal dimension generally does not hold after filtering, which limits the application of fractal-based image model. In this paper, we propose a novel fractal dimension invariant filtering (FDIF) method, extending the invariance of fractal dimension to filtering operations. Utilizing the notion of local self-similarity, we first develop a local fractal model for images. By adding a nonlinear post-processing step behind anisotropic filter banks, we demonstrate that the proposed filtering method is capable of preserving the local invariance of the fractal dimension of image. Meanwhile, we show that the FDIF method can be re-instantiated approximately via a CNN-based architecture, where the convolution layer extracts anisotropic structure of image and the nonlinear layer enhances the structure via preserving local fractal dimension of image. The proposed filtering method provides us with a novel geometric interpretation of CNN-based image model. Focusing on a challenging image processing task - detecting complicated curves from the texture-like images, the proposed method obtains superior results to the state-of-art approaches.

In this work, we explore the learning task of mixtures of Markov chains (MMCs) from aggregate data. Our work demonstrates that although this challenging task is generally intractable because of the identifiability problem, it can be solved approximately by imposing structural constraints on its transition matrices Specifically, the proposed structural constraints include specifying active state sets corresponding to the chains and adding a series of pairwise sparse regularizers on transition matrices. Based on these two structural constraints, we propose a constrained least-squares method to learn mixtures of Markov chains. We develop a novel iterative algorithm that decomposes the overall problem into a set of convex subproblems and solves each subproblem efficiently. Experimental results on synthetic data prove that our learning method converges well and is robust to the noise in data. Moreover, the comparison with state-of-art competitors on real-world data further validates the superiority of our method.

We focus on an important problem of predicting the so-called “patient flow” from longitudinal electronic health records (EHRs), which has not been explored via existing machine learning techniques. We develop a point process based framework for modeling patient flow through various care units (CUs) and jointly predicting patients' destination CUs and duration days. We propose a novel discriminative learning algorithm aiming at improving the prediction of transition events in the case of sparse data. By parameterizing the proposed model as mutually-correcting processes, we formulate the estimation problem via generalized linear models and solve it based on alternating direction method of multipliers (ADMM). We achieve simultaneous feature selection and learning by adding a group-lasso regularizer to the ADMM algorithm. Additionally, we synthesize auxiliary training data for the classes with extremely few samples, and improve the robustness of our learning method to the problem of data imbalance.

Many real-world applications require robust algorithms to learn point process models based on a type of incomplete data — the so-called short doubly-censored (SDC) event sequences. In this paper, we study this critical problem of quantitative asynchronous event sequence analysis under the framework of Hawkes processes by leveraging the general idea of data synthesis. In particular, given SDC event sequences observed in a variety of time intervals, we propose a sampling-stitching data synthesis method — sampling predecessor and successor for each SDC event sequence from potential candidates and stitching them together to synthesize long training sequences. The rationality and the feasibility of our method are discussed in terms of arguments based on likelihood. Experiments on both synthetic and real-world data demonstrate that the proposed data synthesis method improves learning results indeed for both time-invariant and time-varying Hawkes processes.

How to cluster event sequences generated via different point processes is an interesting and important problem in statistical machine learning. To solve this problem, we propose and discuss an effective model-based clustering method based on a novel Dirichlet mixture model of a special but significant type of point processes --- Hawkes process. The proposed model generates the event sequences with different clusters from the Hawkes processes with different parameters, and uses a Dirichlet process as the prior distribution of the clusters. We prove the identifiability of our mixture model and propose an effective variational Bayesian inference algorithm to learn our model. An adaptive inner iteration allocation strategy is designed to accelerate the convergence of our algorithm. Moreover, we investigate the sample complexity and the computational complexity of our learning algorithm in depth. Experiments on both synthetic and real-world data show that the clustering method based on our model can learn structural triggering patterns hidden in asynchronous event sequences robustly and achieve superior performance on clustering purity and consistency compared to existing methods.

Key frames are playing a very important role for many video applications, such as on-line movie preview and video information retrieval. Although a number of key frame selection methods have been proposed in the past, existing technologies mainly focus on how to precisely summarize the video content, but seldom take the user preferences into consideration. However, in real scenarios, people may cast diverse interests on the contents even for the same video, and thus they may be attracted by quite different key frames, which makes the selection of key frames an inherently personalized process. In this paper, we propose and investigate the problem of personalized key frame recommendation to bridge the above gap. To do so, we make use of video images and user time-synchronized comments to design a novel key frame recommender that can simultaneously model visual and textual features in a unified framework. By user personalization based on her/his previously reviewed frames and posted comments, we are able to encode different user interests in a unified multi-modal space, and can thus select key frames in a personalized manner, which, to the best of our knowledge, is the first time in the research field of video content analysis. Experimental results show that our method performs better than its competitors on various measures.

Trajectory analysis is essential in many applications. In this paper, we address the problem of representing motion trajectories in a highly informative way, and consequently utilize it for analyzing trajectories. Our approach first leverages the complete information from given trajectories to construct a thermal transfer field which provides a context-rich way to describe the global motion pattern in a scene. Then, a 3D tube is derived which depicts an input trajectory by integrating its surrounding motion patterns contained in the thermal transfer field. The 3D tube effectively: 1) maintains the movement information of a trajectory, 2) embeds the complete contextual motion pattern around a trajectory, 3) visualizes information about a trajectory in a clear and unified way. We further introduce a droplet-based process. It derives a droplet vector from a 3D tube, so as to characterize the high-dimensional 3D tube information in a simple but effective way. Finally, we apply our tube-and-droplet representation to trajectory analysis applications including trajectory clustering, trajectory classification & abnormality detection, and 3D action recognition. Experimental comparisons with state-of-the-art algorithms demonstrate the effectiveness of our approach.

Over the past decade, the rate of care unit (CU) use in the United States has been increasing. With an aging population and ever-growing demand for medical care, effective management of patients' transitions among different care facilities will prove indispensible for shortening the length of hospital stays, improving patient outcomes, allocating critical care resources, and reducing preventable re-admissions. In this paper, we focus on an important problem of predicting the so-called “patient flow” from longitudinal electronic health records (EHRs), which has not been explored via existing machine learning techniques. By treating a sequence of transition events as a point process, we develop a novel framework for modeling patient flow through various CUs and jointly predicting patients' destination CUs and duration days. Instead of learning a generative point process model via maximum likelihood estimation, we propose a novel discriminative learning algorithm aiming at improving the prediction of transition events in the case of sparse data. By parameterizing the proposed model as a mutually-correcting process, we formulate the estimation problem via generalized linear models, which lends itself to efficient learning based on alternating direction method of multipliers (ADMM). Furthermore, we achieve simultaneous feature selection and learning by adding a group-lasso regularizer to the ADMM algorithm. Additionally, for suppressing the negative influence of data imbalance on the learning of model, we synthesize auxiliary training data for the classes with extremely few samples, and improve the robustness of our learning method accordingly. Testing on real-world data, we show that our method obtains superior performance in terms of accuracy of predicting the destination CU transition and duration of each CU occupancy.

Learning Granger causality for general point processes is a very challenging task. We propose an effective method learning Granger causality for a special but significant type of point processes — Hawkes processes. Focusing on Hawkes processes, we reveal the relationship between Hawkes process’s impact functions and its Granger causality graph. Specifically, our model represents impact functions using a series of basis functions and recovers the Granger causality graph via group sparsity of the impact functions’ coefficients. We propose an effective learning algorithm combining a maximum likelihood estimator (MLE) with a sparse-group-lasso (SGL) regularizer. Additionally, the pairwise similarity between the dimensions of the process is considered when their clustering structure is available. We analyze our learning method and discuss the selection of the basis functions. Experiments on synthetic data and real-world data show that our method can learn the Granger causality graph and the triggering patterns of Hawkes processes simultaneously.

Statistical models based on Markov chains, especially mixtures of Markov chains, have recently been studied and demonstrated to be effective in various data mining applications such as tourist flow analysis, animal migration modeling, and transportation administration. Nevertheless, the research so far has mainly focused on analyzing data at individual levels. Due to security and privacy reasons, however, the observations in practice usually consist of coarse-grained statistics of individual data,a.k.a.aggregate data, rendering learning mixtures of Markov chains an even more challenging problem. In this work, we show that this challenging problem, although intractable in its original form, can be solved approximately by posing structural constraints on the transition matrices. The proposed structural constraints include specifying active state sets corresponding to the chains and adding a pairwise sparse regularization term on transition matrices. Based on these two structural constraints, we propose a constrained least-squares method to learn mixtures of Markov chains. We further develop a novel iterative algorithm that decomposes the overall problem into a set of convex subproblems and solves each subproblem efficiently, making it possible to effectively learn mixtures of Markov chains from aggregate data. We propose a framework for generating synthetic data and analyze the complexity of our algorithm. Additionally, the empirical results of the convergence and the robustness of our algorithm are also presented. These results demonstrate the effectiveness and efficiency of the proposed algorithm, comparing with traditional methods. Experimental results on real-world data sets further validate that our algorithm can be used to solve practical problems.

In this paper, we propose a novel dictionary learning method in the semi-supervised setting by dynamically coupling graph and group structures. To this end, samples are represented by sparse codes inheriting their graph structure while the labeled samples within the same class are represented with group sparsity, sharing the same atoms of the dictionary. Instead of statically combining graph and group structures, we take advantage of them in a mutually reinforcing way — in the dictionary learning phase, we introduce the unlabeled samples into groups by an entropy-based method and then update the corresponding local graph, resulting in a more structured and discriminative dictionary. We analyze the relationship between the two structures and prove the convergence of our proposed method. Focusing on image classification task, we evaluate our approach on several datasets and obtain superior performance compared with the state-of-the-art methods, especially in the case of only a few labeled samples and limited dictionary size.

In this paper, we propose an interpretation of active learning from a pure algebraic view and combine it with semi-supervised manifold learning. The proposed active manifold learning algorithm aims to learn the low-dimensional parameter space of the manifold with high accuracy from smartly labeled samples. We demonstrate that this problem is equivalent to a condition number minimization problem of the alignment matrix. Focusing on this problem, we first give a theoretical upper bound for the solution. Then we develop a heuristic but effective sample selection algorithm with the help of the Gershgorin circle theorem. We investigate the rationality, the feasibility, the universality and the complexity of the proposed method and demonstrate that our method yields encouraging active learning results.

This paper proposes a shrinkage-based framework for unsupervised trajectory clustering. Facing to the challenges of trajectory clustering, e.g., large variations within a cluster and ambiguities across clusters, we first introduce an adaptive multi-kernel-based estimation process to estimate the 'shrunk' positions and speeds of trajectories' points. This kernel-based estimation effectively leverages both multiple structural information within a trajectory and the local motion patterns across multiple trajectories, such that the discrimination of the shrunk point can be properly increased. We further introduce a speed-regularized optimization process, which utilizes the estimated speeds to regularize the optimal shrunk points, so as to guarantee the smoothness and the discriminative pattern of the final shrunk trajectory. Using our approach, the variations among similar trajectories can be reduced while the boundaries between different clusters are enlarged. Experimental results demonstrate that our approach is superior to the state-of-art approaches on both clustering accuracy and robustness. Besides, additional experiments further reveal the effectiveness of our approach when applied to trajectory analysis applications such as anomaly detection and route analysis.

Graph matching has a wide spectrum of real-world applications and in general is known NP-hard. In many vision tasks, one realistic problem arises for finding the global node mappings across a batch of corrupted weighted graphs. This paper is an attempt to connect graph matching, especially multi-graph matching to the matrix decomposition model and its relevant on-the-shelf convex optimization algorithms. Our method aims to extract the common inliers and their synchronized permutations from disordered weighted graphs in the presence of deformation and outliers. Under the proposed framework, several variants can be derived in the hope of accommodating to other types of noises. Experimental results on both synthetic data and real images empirically show that the proposed paradigm exhibits several interesting behaviors and in many cases performs competitively with the state-of-the-arts.

Traditional sparse image models treat color image pixel as a scalar, which represents color channels separately or concatenate color channels as a monochrome image. In this paper, we propose a vector sparse representation model for color images using quaternion matrix analysis. As a new tool for color image representation, its potential applications in several image-processing tasks are presented, including color image reconstruction, denoising, inpainting, and super-resolution. The proposed model represents the color image as a quaternion matrix, where a quaternion-based dictionary learning algorithm is presented using the K-quaternion singular value decomposition (QSVD) (generalized K-means clustering for QSVD) method. It conducts the sparse basis selection in quaternion space, which uniformly transforms the channel images to an orthogonal color space. In this new color space, it is significant that the inherent color structures can be completely preserved during vector reconstruction. Moreover, the proposed sparse model is more efficient comparing with the current sparse models for image restoration tasks due to lower redundancy between the atoms of different color channels. The experimental results demonstrate that the proposed sparse image model avoids the hue bias issue successfully and shows its potential as a general and powerful tool in color image analysis and processing domain.

In this paper, we present a novel method to synthesize dynamic texture sequences from extremely few samples, e.g., merely two possibly disparate frames, leveraging both Markov Random Fields (MRFs) and manifold learning. Decomposing a textural image as a set of patches, we achieve dynamic texture synthesis by estimating sequences of temporal patches. We select candidates for each temporal patch from spatial patches based on MRFs and regard them as samples from a low-dimensional manifold. After mapping candidates to a low-dimensional latent space, we estimate the sequence of temporal patches by finding an optimal trajectory in the latent space. Guided by some key properties of trajectories of realistic temporal patches, we derive a curvature-based trajectory selection algorithm. In contrast to the methods based on MRFs or dynamic systems that rely on a large amount of samples, our method is able to deal with the case of extremely few samples and requires no training phase. We compare our method with the state of the art and show that our method not only exhibits superior performance on synthesizing textures but it also produces results with pleasing visual effects.

In this paper, we propose a generalized equalization model for image enhancement. Based on our analysis on the relationships between image histogram and contrast enhancement/white balancing, we first establish a generalized equalization model integrating contrast enhancement and white balancing into a unified framework of convex programming of image histogram. We show that many image enhancement tasks can be accomplished by the proposed model using different configurations of parameters. With two defining properties of histogram transform, namely contrast gain and nonlinearity, the model parameters for different enhancement applications can be optimized. We then derive an optimal image enhancement algorithm that theoretically achieves the best joint contrast enhancement and white balancing result with trading-off between contrast enhancement and tonal distortion. Subjective and objective experimental results show favorable performances of the proposed algorithm in applications of image enhancement, white balancing and tone correction. Computational complexity of the proposed method is also analyzed.

Data sparsity has been a thorny issue for manifold-based image synthesis, and in this paper we address this critical problem by leveraging ideas from transfer learning. Specifically, we propose methods based on generating auxiliary data in the form of synthetic samples using transformations of the original sparse samples. To incorporate the auxiliary data, we propose a weighted data synthesis method, which adaptively selects from the generated samples for inclusion during the manifold learning process via a weighted iterative algorithm. To demonstrate the feasibility of the proposed method, we apply it to the problem of face image synthesis from sparse samples. Compared with existing methods, the proposed method shows encouraging results with good performance improvements.

In this paper, we propose a quaternion-based sparse representation model for color images and its corresponding dictionary learning algorithm. Differing from traditional sparse image models, which represent RGB channels separately or process RGB channels as a concatenated real vector, the proposed model describes the color image as a quaternion vector matrix, where each color pixel is encoded as a quaternion unit and thus the inter-relationship among RGB channels is well preserved. Correspondingly, we propose a quaternion-based dictionary learning algorithm using a socalled K-QSVD method. It conducts the sparse basis selection in quaternion vector space, providing a kind of vectorial representation for the inherent color structures rather than a scalar representation via current sparse image models. The proposed sparse model is validated in the applications of color image denoising and inpainting. The experimental results demonstrate that our sparse image model avoids the hue bias phenomenon successfully and shows its potential as a powerful tool in color image analysis and processing domain.

In this paper, we propose a single image super-resolution and enhancement algorithm using local fractal analysis. If we treat the pixels of a natural image as a fractal set, the image gradient can then be regarded as a measure of the fractal set. According to the scale invariance (a special case of bi-Lipschitz invariance) feature of fractal dimension, we will be able to estimate the gradient of a high-resolution image from that of a low-resolution one. Moreover, the high-resolution image can be further enhanced by preserving the local fractal length of gradient during the up-sampling process. We show that a regularization term based on the scale invariance of fractal dimension and length can be effective in recovering details of the high-resolution image. Analysis is provided on the relation and difference among the proposed approach and some other state of the art interpolation methods. Experimental results show that the proposed method has superior super-resolution and enhancement results as compared to other competitors.

In this paper, a new ECG signal compression algorithm based on wave atom transform is presented. According to an assumption that ECG is oscillatory signal, we decompose ECG signal by wave atoms and trimmed insignificant coefficients. The wave atom decomposition has been proved to have a significantly sparser solution than other existing transform methods when it comes to oscillatory signal. In our experiment, the convergence of the energy of wave atoms' coefficients is faster than that of wavelet indeed. The most significant advantage of our algorithm is that unlike many conventional methods, the performance of our algorithm is not dependent on QRS detection, which simplifies the architecture of compression system and is beneficial to telemedicine application. After wave atom transform, the data stream is divided and coded by a hybrid entropy coding strategy combining delta coding, run-length-coding and arithmetic coding. The experimental results on MIT-BIH arrhythmia database proved that our algorithm has high compression ratio (CR >; 10) with percentage root mean square difference (PRD) under 1%.