Transfer learning is a cornerstone of modern deep learning, yet it remains constrained by challenges in model selection and the overhead of extensive model storage. In this work, we present Diffusion-based Neural Network Weights Generation, D2NWG, a novel framework that leverages diffusion processes to synthesize task-specific network weights. By modeling the distribution of weights from a diverse ensemble of pretrained models and conditioning the generation process on dataset characteristics, task descriptions, and architectural specifications, D2NWG circumvents the need for storing and searching through massive model repositories. We evaluate D2NWG across multiple experimental settings. On in-distribution tasks, our framework achieves performance that is on par with or superior to conventional pretrained models, while also serving as an effective initialization strategy for novel domains, resulting in faster convergence and a 6\% improvement in few-shot learning scenarios. Extensive ablation studies further indicate that our approach scales robustly with increased diversity and volume of pretrained models. Moreover, D2NWG demonstrates significant promise for large language model applications. In evaluations on the OpenLM leaderboard, our method improved LLaMA-3-2-1B-Instruct performance by 3\% on challenging mathematical reasoning tasks, with a consistent gain of 0.36\% across a range of benchmarks. These findings establish D2NWG as a versatile and powerful framework for neural network weight generation, offering a scalable solution to the limitations of traditional transfer learning.

Anomaly detection is essential for image-based industrial inspection, yet class-specific models hinder its scalability and adaptability. This issue is exacerbated by ‘normality confusion’, where models struggle to distinguish normal from abnormal features across diverse classes. To address this, we introduce TwoStudents, a unified anomaly detection framework utilizing a knowledge distillation model with dual student decoders for efficient feature disentanglement. By separating class-specific abnormal features from normal ones, TwoStudents strengthens the unified normal representation. Our approach employs image transformations to simulate anomalies and uses cross-distillation training techniques on features derived from normal images and their corresponding simulated abnormal images. This effectively isolates abnormal features, enabling a unified model applicable to all classes. Evaluations on the MVTec AD, VisA, and BTAD datasets demonstrate that TwoStudents outperforms existing methods, achieving state-of-the-art performance. Notably, it significantly improves fine-grained localization, as measured by pixel AUPRO, with increases from 91.9% to 94.0% on MVTec AD and 87.6% to 90.5% on VisA.

Existing NAS methods suffer from either an excessive amount of time for repetitive sampling and training of many task-irrelevant architectures. To tackle such limitations of existing NAS methods, we propose a paradigm shift from NAS to a novel conditional Neural Architecture Generation (NAG) framework based on diffusion models, dubbed DiffusionNAG. Specifically, we consider the neural architectures as directed graphs and propose a graph diffusion model for generating them. Moreover, with the guidance of parameterized predictors, DiffusionNAG can flexibly generate task-optimal architectures with the desired properties for diverse tasks, by sampling from a region that is more likely to satisfy the properties. This conditional NAG scheme is significantly more efficient than previous NAS schemes which sample the architectures and filter them using the property predictors. We validate the effectiveness of DiffusionNAG through extensive experiments in two predictor-based NAS scenarios: Transferable NAS and Bayesian Optimization (BO)-based NAS. DiffusionNAG achieves superior performance with speedups of up to 35$\times$ when compared to the baselines on Transferable NAS benchmarks. Furthermore, when integrated into a BO-based algorithm, DiffusionNAG outperforms existing BO-based NAS approaches, particularly in the large MobileNetV3 search space on the ImageNet 1K dataset. Code is available at https://github.com/CownowAn/DiffusionNAG.

Being able to use a single app across multiple devices can bring novel experiences to the users in various domains including entertainment and productivity. For instance, a user of a video editing app would be able to use a smart pad as a canvas and a smartphone as a remote toolbox so that the toolbox does not occlude the canvas during editing. However, existing approaches do not properly support the single-app multi-device execution due to several limitations, including high development cost, device heterogeneity, and high performance requirement. In this paper, we introduce FLUID-XP, a novel cross-platform multi-device system that enables UIs of a single app to be executed across heterogeneous platforms, while overcoming the limitations of previous approaches. FLUID-XP provides flexible, efficient, and seamless interactions by addressing three main challenges: i) how to transparently enable a single-display app to use multiple displays, ii) how to distribute UIs across heterogeneous devices with minimal network traffic, and iii) how to optimize the UI distribution process when multiple UIs have different distribution requirements. Our experiments with a working prototype of FLUID-XP on Android confirm that FLUID-XP successfully supports a variety of unmodified real-world apps across heterogeneous platforms (Android, iOS, and Linux). We also conduct a lab study with 25 participants to demonstrate the effectiveness of FLUID-XP with real users.

Most conventional Neural Architecture Search (NAS) approaches are limited in that they only generate architectures without searching for the optimal parameters. While some NAS methods handle this issue by utilizing a supernet trained on a large-scale dataset such as ImageNet, they may be suboptimal if the target tasks are highly dissimilar from the dataset the supernet is trained on. To address such limitations, we introduce a novel problem of Neural Network Search (NNS), whose goal is to search for the optimal pretrained network for a novel dataset and constraints (e.g. number of parameters), from a model zoo. Then, we propose a novel framework to tackle the problem, namely Task-Adaptive Neural Network Search (TANS). Given a model-zoo that consists of network pretrained on diverse datasets, we use a novel amortized meta-learning framework to learn a cross-modal latent space with contrastive loss, to maximize the similarity between a dataset and a high-performing network on it, and minimize the similarity between irrelevant dataset-network pairs. We validate the effectiveness and efficiency of our method on ten real-world datasets, against existing NAS/AutoML baselines. The results show that our method instantly retrieves networks that outperform models obtained with the baselines with significantly fewer training steps to reach the target performance, thus minimizing the total cost of obtaining a task-optimal network. Our code and the model-zoo are available at https://anonymous.4open.science/r/TANS-33D6

For deployment, neural architecture search should be hardware-aware, in order to satisfy the device-specific constraints (e.g., memory usage, latency and energy consumption) and enhance the model efficiency. Existing methods on hardware-aware NAS collect a large number of samples (e.g., accuracy and latency) from a target device, either builds a lookup table or a latency estimator. However, such approach is impractical in real-world scenarios as there exist numerous devices with different hardware specifications, and collecting samples from such a large number of devices will require prohibitive computational and monetary cost. To overcome such limitations, we propose Hardware-adaptive Efficient Latency Predictor (HELP), which formulates the device-specific latency estimation problem as a meta-learning problem, such that we can estimate the latency of a model's performance for a given task on an unseen device with a few samples. To this end, we introduce novel hardware embeddings to embed any devices considering them as black-box functions that output latencies, and meta-learn the hardware-adaptive latency predictor in a device-dependent manner, using the hardware embeddings. We validate the proposed HELP for its latency estimation performance on unseen platforms, on which it achieves high estimation performance with as few as 10 measurement samples, outperforming all relevant baselines. We also validate end-to-end NAS frameworks using HELP against ones without it, and show that it largely reduces the total time cost of the base NAS method, in latency-constrained settings.