WHALE: Towards Generalizable and Scalable World Models for Embodied Decision-making

World models play a crucial role in decision-making within embodied environments, enabling cost-free explorations that would otherwise be expensive in the real world. To facilitate effective decision-making, world models must be equipped with strong generalizability to support faithful imagination in out-of-distribution (OOD) regions and provide reliable uncertainty estimation to assess the credibility of the simulated experiences, both of which present significant challenges for prior scalable approaches. This paper introduces WHALE, a framework for learning generalizable world models, consisting of two key techniques: behavior-conditioning and retracing-rollout. Behavior-conditioning addresses the policy distribution shift, one of the primary sources of the world model generalization error, while retracing-rollout enables efficient uncertainty estimation without the necessity of model ensembles. These techniques are universal and can be combined with any neural network architecture for world model learning. Incorporating these two techniques, we present Whale-ST, a scalable spatial-temporal transformer-based world model with enhanced generalizability. We demonstrate the superiority of Whale-ST in simulation tasks by evaluating both value estimation accuracy and video generation fidelity. Additionally, we examine the effectiveness of our uncertainty estimation technique, which enhances model-based policy optimization in fully offline scenarios. Furthermore, we propose Whale-X, a 414M parameter world model trained on 970K trajectories from Open X-Embodiment datasets. We show that Whale-X exhibits promising scalability and strong generalizability in real-world manipulation scenarios using minimal demonstrations.

Provably and Practically Efficient Adversarial Imitation Learning with General Function Approximation

As a prominent category of imitation learning methods, adversarial imitation learning (AIL) has garnered significant practical success powered by neural network approximation. However, existing theoretical studies on AIL are primarily limited to simplified scenarios such as tabular and linear function approximation and involve complex algorithmic designs that hinder practical implementation, highlighting a gap between theory and practice. In this paper, we explore the theoretical underpinnings of online AIL with general function approximation. We introduce a new method called optimization-based AIL (OPT-AIL), which centers on performing online optimization for reward functions and optimism-regularized Bellman error minimization for Q-value functions. Theoretically, we prove that OPT-AIL achieves polynomial expert sample complexity and interaction complexity for learning near-expert policies. To our best knowledge, OPT-AIL is the first provably efficient AIL method with general function approximation. Practically, OPT-AIL only requires the approximate optimization of two objectives, thereby facilitating practical implementation. Empirical studies demonstrate that OPT-AIL outperforms previous state-of-the-art deep AIL methods in several challenging tasks.

MPT: A Large-scale Multi-Phytoplankton Tracking Benchmark

Phytoplankton are a crucial component of aquatic ecosystems, and effective monitoring of them can provide valuable insights into ocean environments and ecosystem changes. Traditional phytoplankton monitoring methods are often complex and lack timely analysis. Therefore, deep learning algorithms offer a promising approach for automated phytoplankton monitoring. However, the lack of large-scale, high-quality training samples has become a major bottleneck in advancing phytoplankton tracking. In this paper, we propose a challenging benchmark dataset, Multiple Phytoplankton Tracking (MPT), which covers diverse background information and variations in motion during observation. The dataset includes 27 species of phytoplankton and zooplankton, 14 different backgrounds to simulate diverse and complex underwater environments, and a total of 140 videos. To enable accurate real-time observation of phytoplankton, we introduce a multi-object tracking method, Deviation-Corrected Multi-Scale Feature Fusion Tracker(DSFT), which addresses issues such as focus shifts during tracking and the loss of small target information when computing frame-to-frame similarity. Specifically, we introduce an additional feature extractor to predict the residuals of the standard feature extractor's output, and compute multi-scale frame-to-frame similarity based on features from different layers of the extractor. Extensive experiments on the MPT have demonstrated the validity of the dataset and the superiority of DSFT in tracking phytoplankton, providing an effective solution for phytoplankton monitoring.

LLMOPT: Learning to Define and Solve General Optimization Problems from Scratch

Optimization problems are prevalent across various scenarios. Formulating and then solving optimization problems described by natural language often requires highly specialized human expertise, which could block the widespread application of optimization-based decision making. To make problem formulating and solving automated, leveraging large language models (LLMs) has emerged as a potential way. However, this kind of way suffers from the issue of optimization generalization. Namely, the accuracy of most current LLM-based methods and the generality of optimization problem types that they can model are still limited. In this paper, we propose a unified learning-based framework called LLMOPT to boost optimization generalization. Starting from the natural language descriptions of optimization problems and a pre-trained LLM, LLMOPT constructs the introduced five-element formulation as a universal model for learning to define diverse optimization problem types. Then, LLMOPT employs the multi-instruction tuning to enhance both problem formalization and solver code generation accuracy and generality. After that, to prevent hallucinations in LLMs, such as sacrificing solving accuracy to avoid execution errors, model alignment and self-correction mechanism are adopted in LLMOPT. We evaluate the optimization generalization ability of LLMOPT and compared methods across six real-world datasets covering roughly 20 fields such as health, environment, energy and manufacturing, etc. Extensive experiment results show that LLMOPT is able to model various optimization problem types such as linear/nonlinear programming, mixed integer programming and combinatorial optimization, and achieves a notable 11.08% average solving accuracy improvement compared with the state-of-the-art methods. The code is available at https://github.com/caigaojiang/LLMOPT.

Dual-AEB: Synergizing Rule-Based and Multimodal Large Language Models for Effective Emergency Braking

Automatic Emergency Braking (AEB) systems are a crucial component in ensuring the safety of passengers in autonomous vehicles. Conventional AEB systems primarily rely on closed-set perception modules to recognize traffic conditions and assess collision risks. To enhance the adaptability of AEB systems in open scenarios, we propose Dual-AEB, a system combines an advanced multimodal large language model (MLLM) for comprehensive scene understanding and a conventional rule-based rapid AEB to ensure quick response times. To the best of our knowledge, Dual-AEB is the first method to incorporate MLLMs within AEB systems. Through extensive experimentation, we have validated the effectiveness of our method. The source code will be available at https://github.com/ChipsICU/Dual-AEB.

ZipGait: Bridging Skeleton and Silhouette with Diffusion Model for Advancing Gait Recognition

Current gait recognition research predominantly focuses on extracting appearance features effectively, but the performance is severely compromised by the vulnerability of silhouettes under unconstrained scenes. Consequently, numerous studies have explored how to harness information from various models, particularly by sufficiently utilizing the intrinsic information of skeleton sequences. While these model-based methods have achieved significant performance, there is still a huge gap compared to appearance-based methods, which implies the potential value of bridging silhouettes and skeletons. In this work, we make the first attempt to reconstruct dense body shapes from discrete skeleton distributions via the diffusion model, demonstrating a new approach that connects cross-modal features rather than focusing solely on intrinsic features to improve model-based methods. To realize this idea, we propose a novel gait diffusion model named DiffGait, which has been designed with four specific adaptations suitable for gait recognition. Furthermore, to effectively utilize the reconstructed silhouettes and skeletons, we introduce Perception Gait Integration (PGI) to integrate different gait features through a two-stage process. Incorporating those modifications leads to an efficient model-based gait recognition framework called ZipGait. Through extensive experiments on four public benchmarks, ZipGait demonstrates superior performance, outperforming the state-of-the-art methods by a large margin under both cross-domain and intra-domain settings, while achieving significant plug-and-play performance improvements.

TS-SAM: Fine-Tuning Segment-Anything Model for Downstream Tasks

Adapter based fine-tuning has been studied for improving the performance of SAM on downstream tasks. However, there is still a significant performance gap between fine-tuned SAMs and domain-specific models. To reduce the gap, we propose Two-Stream SAM (TS-SAM). On the one hand, inspired by the side network in Parameter-Efficient Fine-Tuning (PEFT), we designed a lightweight Convolutional Side Adapter (CSA), which integrates the powerful features from SAM into side network training for comprehensive feature fusion. On the other hand, in line with the characteristics of segmentation tasks, we designed Multi-scale Refinement Module (MRM) and Feature Fusion Decoder (FFD) to keep both the detailed and semantic features. Extensive experiments on ten public datasets from three tasks demonstrate that TS-SAM not only significantly outperforms the recently proposed SAM-Adapter and SSOM, but achieves competitive performance with the SOTA domain-specific models. Our code is available at: https://github.com/maoyangou147/TS-SAM.

Energy-Guided Diffusion Sampling for Offline-to-Online Reinforcement Learning

Combining offline and online reinforcement learning (RL) techniques is indeed crucial for achieving efficient and safe learning where data acquisition is expensive. Existing methods replay offline data directly in the online phase, resulting in a significant challenge of data distribution shift and subsequently causing inefficiency in online fine-tuning. To address this issue, we introduce an innovative approach, \textbf{E}nergy-guided \textbf{DI}ffusion \textbf{S}ampling (EDIS), which utilizes a diffusion model to extract prior knowledge from the offline dataset and employs energy functions to distill this knowledge for enhanced data generation in the online phase. The theoretical analysis demonstrates that EDIS exhibits reduced suboptimality compared to solely utilizing online data or directly reusing offline data. EDIS is a plug-in approach and can be combined with existing methods in offline-to-online RL setting. By implementing EDIS to off-the-shelf methods Cal-QL and IQL, we observe a notable 20% average improvement in empirical performance on MuJoCo, AntMaze, and Adroit environments. Code is available at \url{https://github.com/liuxhym/EDIS}.

UnmixingSR: Material-aware Network with Unsupervised Unmixing as Auxiliary Task for Hyperspectral Image Super-resolution

Deep learning-based (DL-based) hyperspectral image (HIS) super-resolution (SR) methods have achieved remarkable performance and attracted attention in industry and academia. Nonetheless, most current methods explored and learned the mapping relationship between low-resolution (LR) and high-resolution (HR) HSIs, leading to the side effect of increasing unreliability and irrationality in solving the ill-posed SR problem. We find, quite interestingly, LR imaging is similar to the mixed pixel phenomenon. A single photodetector in sensor arrays receives the reflectance signals reflected by a number of classes, resulting in low spatial resolution and mixed pixel problems. Inspired by this observation, this paper proposes a component-aware HSI SR network called UnmixingSR, in which the unsupervised HU as an auxiliary task is used to perceive the material components of HSIs. We regard HU as an auxiliary task and incorporate it into the HSI SR process by exploring the constraints between LR and HR abundances. Instead of only learning the mapping relationship between LR and HR HSIs, we leverage the bond between LR abundances and HR abundances to boost the stability of our method in solving SR problems. Moreover, the proposed unmixing process can be embedded into existing deep SR models as a plug-in-play auxiliary task. Experimental results on hyperspectral experiments show that unmixing process as an auxiliary task incorporated into the SR problem is feasible and rational, achieving outstanding performance. The code is available at

Hindsight Preference Learning for Offline Preference-based Reinforcement Learning

Offline preference-based reinforcement learning (RL), which focuses on optimizing policies using human preferences between pairs of trajectory segments selected from an offline dataset, has emerged as a practical avenue for RL applications. Existing works rely on extracting step-wise reward signals from trajectory-wise preference annotations, assuming that preferences correlate with the cumulative Markovian rewards. However, such methods fail to capture the holistic perspective of data annotation: Humans often assess the desirability of a sequence of actions by considering the overall outcome rather than the immediate rewards. To address this challenge, we propose to model human preferences using rewards conditioned on future outcomes of the trajectory segments, i.e. the hindsight information. For downstream RL optimization, the reward of each step is calculated by marginalizing over possible future outcomes, the distribution of which is approximated by a variational auto-encoder trained using the offline dataset. Our proposed method, Hindsight Preference Learning (HPL), can facilitate credit assignment by taking full advantage of vast trajectory data available in massive unlabeled datasets. Comprehensive empirical studies demonstrate the benefits of HPL in delivering robust and advantageous rewards across various domains. Our code is publicly released at https://github.com/typoverflow/WiseRL.