Ye Shi

Hongxia Li, Zhongyi Cai, Jingya Wang et al.

Federated learning is an emerging learning paradigm where multiple clients collaboratively train a machine learning model in a privacy-preserving manner. Personalized federated learning extends this paradigm to overcome heterogeneity across clients by learning personalized models. Recently, there have been some initial attempts to apply Transformers to federated learning. However, the impacts of federated learning algorithms on self-attention have not yet been studied. This paper investigates this relationship and reveals that federated averaging algorithms actually have a negative impact on self-attention where there is data heterogeneity. These impacts limit the capabilities of the Transformer model in federated learning settings. Based on this, we propose FedTP, a novel Transformer-based federated learning framework that learns personalized self-attention for each client while aggregating the other parameters among the clients. Instead of using a vanilla personalization mechanism that maintains personalized self-attention layers of each client locally, we develop a learn-to-personalize mechanism to further encourage the cooperation among clients and to increase the scablability and generalization of FedTP. Specifically, the learn-to-personalize is realized by learning a hypernetwork on the server that outputs the personalized projection matrices of self-attention layers to generate client-wise queries, keys and values. Furthermore, we present the generalization bound for FedTP with the learn-to-personalize mechanism. Notably, FedTP offers a convenient environment for performing a range of image and language tasks using the same federated network architecture - all of which benefit from Transformer personalization. Extensive experiments verify that FedTP with the learn-to-personalize mechanism yields state-of-the-art performance in non-IID scenarios. Our code is available online.

Haixiang Sun, Ye Shi, Jingya Wang et al.

The idea of embedding optimization problems into deep neural networks as optimization layers to encode constraints and inductive priors has taken hold in recent years. Most existing methods focus on implicitly differentiating Karush-Kuhn-Tucker (KKT) conditions in a way that requires expensive computations on the Jacobian matrix, which can be slow and memory-intensive. In this paper, we developed a new framework, named Alternating Differentiation (Alt-Diff), that differentiates optimization problems (here, specifically in the form of convex optimization problems with polyhedral constraints) in a fast and recursive way. Alt-Diff decouples the differentiation procedure into a primal update and a dual update in an alternating way. Accordingly, Alt-Diff substantially decreases the dimensions of the Jacobian matrix especially for optimization with large-scale constraints and thus increases the computational speed of implicit differentiation. We show that the gradients obtained by Alt-Diff are consistent with those obtained by differentiating KKT conditions. In addition, we propose to truncate Alt-Diff to further accelerate the computational speed. Under some standard assumptions, we show that the truncation error of gradients is upper bounded by the same order of variables' estimation error. Therefore, Alt-Diff can be truncated to further increase computational speed without sacrificing much accuracy. A series of comprehensive experiments validate the superiority of Alt-Diff.

Wanxing Chang, Ye Shi, Hoang Duong Tuan et al.

Universal Domain Adaptation (UniDA) aims to transfer knowledge from a source domain to a target domain without any constraints on label sets. Since both domains may hold private classes, identifying target common samples for domain alignment is an essential issue in UniDA. Most existing methods require manually specified or hand-tuned threshold values to detect common samples thus they are hard to extend to more realistic UniDA because of the diverse ratios of common classes. Moreover, they cannot recognize different categories among target-private samples as these private samples are treated as a whole. In this paper, we propose to use Optimal Transport (OT) to handle these issues under a unified framework, namely UniOT. First, an OT-based partial alignment with adaptive filling is designed to detect common classes without any predefined threshold values for realistic UniDA. It can automatically discover the intrinsic difference between common and private classes based on the statistical information of the assignment matrix obtained from OT. Second, we propose an OT-based target representation learning that encourages both global discrimination and local consistency of samples to avoid the over-reliance on the source. Notably, UniOT is the first method with the capability to automatically discover and recognize private categories in the target domain for UniDA. Accordingly, we introduce a new metric H^3-score to evaluate the performance in terms of both accuracy of common samples and clustering performance of private ones. Extensive experiments clearly demonstrate the advantages of UniOT over a wide range of state-of-the-art methods in UniDA.

Yachen Gao, Xinwei Sun, Yikai Wang et al.

Conditional generative models have recently achieved remarkable success in various applications. However, a suitable metric for evaluating the reliability of these models, which takes into account their inherent uncertainty, is still lacking. Existing metrics, which typically assess a single output, may fail to capture the variability or potential risks in generation. In this paper, we propose a novel evaluation metric called reliability score based on conformal prediction, which measures the worst-case performance within the prediction set at a pre-specified confidence level. However, computing this score is challenging due to the high-dimensional nature of the output space and the nonconvexity of both the metric function and the prediction set. To efficiently compute this score, we introduce Conformal ReLiability (CReL), a framework that can (i) construct the prediction set with desired coverage; and (ii) accurately optimize the reliability score within the constructed prediction set. We provide theoretical results on coverage and demonstrate empirically that our method produces more informative prediction sets than existing approaches. Experiments on synthetic data and the image-to-text and text-to-image tasks further demonstrate the interpretability of our new metric, and the validity and effectiveness of our computational framework. Source code can be found at https://ggc29.github.io/CReL/.

Chunlin Yu, Ye Shi, Zimo Liu et al.

Lifelong person re-identification (LReID) is in significant demand for real-world development as a large amount of ReID data is captured from diverse locations over time and cannot be accessed at once inherently. However, a key challenge for LReID is how to incrementally preserve old knowledge and gradually add new capabilities to the system. Unlike most existing LReID methods, which mainly focus on dealing with catastrophic forgetting, our focus is on a more challenging problem, which is, not only trying to reduce the forgetting on old tasks but also aiming to improve the model performance on both new and old tasks during the lifelong learning process. Inspired by the biological process of human cognition where the somatosensory neocortex and the hippocampus work together in memory consolidation, we formulated a model called Knowledge Refreshing and Consolidation (KRC) that achieves both positive forward and backward transfer. More specifically, a knowledge refreshing scheme is incorporated with the knowledge rehearsal mechanism to enable bi-directional knowledge transfer by introducing a dynamic memory model and an adaptive working model. Moreover, a knowledge consolidation scheme operating on the dual space further improves model stability over the long term. Extensive evaluations show KRC's superiority over the state-of-the-art LReID methods on challenging pedestrian benchmarks.

Zixuan Zhang, Yuning Jiang, Yuanming Shi et al.

With the recent advances in mobile energy storage technologies, electric vehicles (EVs) have become a crucial part of smart grids. When EVs participate in the demand response program, the charging cost can be significantly reduced by taking full advantage of the real-time pricing signals. However, many stochastic factors exist in the dynamic environment, bringing significant challenges to design an optimal charging/discharging control strategy. This paper develops an optimal EV charging/discharging control strategy for different EV users under dynamic environments to maximize EV users' benefits. We first formulate this problem as a Markov decision process (MDP). Then we consider EV users with different behaviors as agents in different environments. Furthermore, a horizontal federated reinforcement learning (HFRL)-based method is proposed to fit various users' behaviors and dynamic environments. This approach can learn an optimal charging/discharging control strategy without sharing users' profiles. Simulation results illustrate that the proposed real-time EV charging/discharging control strategy can perform well among various stochastic factors.

Juze Zhang, Haimin Luo, Hongdi Yang et al.

Humans constantly interact with objects in daily life tasks. Capturing such processes and subsequently conducting visual inferences from a fixed viewpoint suffers from occlusions, shape and texture ambiguities, motions, etc. To mitigate the problem, it is essential to build a training dataset that captures free-viewpoint interactions. We construct a dense multi-view dome to acquire a complex human object interaction dataset, named HODome, that consists of $\sim$75M frames on 10 subjects interacting with 23 objects. To process the HODome dataset, we develop NeuralDome, a layer-wise neural processing pipeline tailored for multi-view video inputs to conduct accurate tracking, geometry reconstruction and free-view rendering, for both human subjects and objects. Extensive experiments on the HODome dataset demonstrate the effectiveness of NeuralDome on a variety of inference, modeling, and rendering tasks. Both the dataset and the NeuralDome tools will be disseminated to the community for further development.

Leijie Zhang, Ye Shi, Yu-Cheng Chang et al.

Heterogeneous big data poses many challenges in machine learning. Its enormous scale, high dimensionality, and inherent uncertainty make almost every aspect of machine learning difficult, from providing enough processing power to maintaining model accuracy to protecting privacy. However, perhaps the most imposing problem is that big data is often interspersed with sensitive personal data. Hence, we propose a privacy-preserving hierarchical fuzzy neural network (PP-HFNN) to address these technical challenges while also alleviating privacy concerns. The network is trained with a two-stage optimization algorithm, and the parameters at low levels of the hierarchy are learned with a scheme based on the well-known alternating direction method of multipliers, which does not reveal local data to other agents. Coordination at high levels of the hierarchy is handled by the alternating optimization method, which converges very quickly. The entire training procedure is scalable, fast and does not suffer from gradient vanishing problems like the methods based on back-propagation. Comprehensive simulations conducted on both regression and classification tasks demonstrate the effectiveness of the proposed model.

Yu-Tong Cao, Ye Shi, Baosheng Yu et al.

Federated learning enables multiple decentralized clients to learn collaboratively without sharing the local training data. However, the expensive annotation cost to acquire data labels on local clients remains an obstacle in utilizing local data. In this paper, we propose a federated active learning paradigm to efficiently learn a global model with limited annotation budget while protecting data privacy in a decentralized learning way. The main challenge faced by federated active learning is the mismatch between the active sampling goal of the global model on the server and that of the asynchronous local clients. This becomes even more significant when data is distributed non-IID across local clients. To address the aforementioned challenge, we propose Knowledge-Aware Federated Active Learning (KAFAL), which consists of Knowledge-Specialized Active Sampling (KSAS) and Knowledge-Compensatory Federated Update (KCFU). KSAS is a novel active sampling method tailored for the federated active learning problem. It deals with the mismatch challenge by sampling actively based on the discrepancies between local and global models. KSAS intensifies specialized knowledge in local clients, ensuring the sampled data to be informative for both the local clients and the global model. KCFU, in the meantime, deals with the client heterogeneity caused by limited data and non-IID data distributions. It compensates for each client's ability in weak classes by the assistance of the global model. Extensive experiments and analyses are conducted to show the superiority of KSAS over the state-of-the-art active learning methods and the efficiency of KCFU under the federated active learning framework.

Juze Zhang, Ye Shi, Yuexin Ma et al.

This paper presents an inverse kinematic optimization layer (IKOL) for 3D human pose and shape estimation that leverages the strength of both optimization- and regression-based methods within an end-to-end framework. IKOL involves a nonconvex optimization that establishes an implicit mapping from an image's 3D keypoints and body shapes to the relative body-part rotations. The 3D keypoints and the body shapes are the inputs and the relative body-part rotations are the solutions. However, this procedure is implicit and hard to make differentiable. So, to overcome this issue, we designed a Gauss-Newton differentiation (GN-Diff) procedure to differentiate IKOL. GN-Diff iteratively linearizes the nonconvex objective function to obtain Gauss-Newton directions with closed form solutions. Then, an automatic differentiation procedure is directly applied to generate a Jacobian matrix for end-to-end training. Notably, the GN-Diff procedure works fast because it does not rely on a time-consuming implicit differentiation procedure. The twist rotation and shape parameters are learned from the neural networks and, as a result, IKOL has a much lower computational overhead than most existing optimization-based methods. Additionally, compared to existing regression-based methods, IKOL provides a more accurate mesh-image correspondence. This is because it iteratively reduces the distance between the keypoints and also enhances the reliability of the pose structures. Extensive experiments demonstrate the superiority of our proposed framework over a wide range of 3D human pose and shape estimation methods.

Juze Zhang, Jingya Wang, Ye Shi et al.

Inter-person occlusion and depth ambiguity make estimating the 3D poses of monocular multiple persons as camera-centric coordinates a challenging problem. Typical top-down frameworks suffer from high computational redundancy with an additional detection stage. By contrast, the bottom-up methods enjoy low computational costs as they are less affected by the number of humans. However, most existing bottom-up methods treat camera-centric 3D human pose estimation as two unrelated subtasks: 2.5D pose estimation and camera-centric depth estimation. In this paper, we propose a unified model that leverages the mutual benefits of both these subtasks. Within the framework, a robust structured 2.5D pose estimation is designed to recognize inter-person occlusion based on depth relationships. Additionally, we develop an end-to-end geometry-aware depth reasoning method that exploits the mutual benefits of both 2.5D pose and camera-centric root depths. This method first uses 2.5D pose and geometry information to infer camera-centric root depths in a forward pass, and then exploits the root depths to further improve representation learning of 2.5D pose estimation in a backward pass. Further, we designed an adaptive fusion scheme that leverages both visual perception and body geometry to alleviate inherent depth ambiguity issues. Extensive experiments demonstrate the superiority of our proposed model over a wide range of bottom-up methods. Our accuracy is even competitive with top-down counterparts. Notably, our model runs much faster than existing bottom-up and top-down methods.

Ye Shi, Leijie Zhang, Zehong Cao et al.

Recently, distributed semi-supervised learning (DSSL) algorithms have shown their effectiveness in leveraging unlabeled samples over interconnected networks, where agents cannot share their original data with each other and can only communicate non-sensitive information with their neighbors. However, existing DSSL algorithms cannot cope with data uncertainties and may suffer from high computation and communication overhead problems. To handle these issues, we propose a distributed semi-supervised fuzzy regression (DSFR) model with fuzzy if-then rules and interpolation consistency regularization (ICR). The ICR, which was proposed recently for semi-supervised problem, can force decision boundaries to pass through sparse data areas, thus increasing model robustness. However, its application in distributed scenarios has not been considered yet. In this work, we proposed a distributed Fuzzy C-means (DFCM) method and a distributed interpolation consistency regularization (DICR) built on the well-known alternating direction method of multipliers to respectively locate parameters in antecedent and consequent components of DSFR. Notably, the DSFR model converges very fast since it does not involve back-propagation procedure and is scalable to large-scale datasets benefiting from the utilization of DFCM and DICR. Experiments results on both artificial and real-world datasets show that the proposed DSFR model can achieve much better performance than the state-of-the-art DSSL algorithm in terms of both loss value and computational cost.

Shutong Ding, Zejia Zhong, Zhongyi Wang et al.

Recent advances in reinforcement learning (RL) have achieved great successes by leveraging the multimodality and exploration capability of diffusion policies. Among these approaches, one representative branch focuses on the sampling-based policy optimization. This design enables better exploration capability of the diffusion model, particularly at the beginning of training, but suffer from low exploitation in Q-value information, resulting in a slow policy convergence. Another branch pays attention to gradient-based policy optimization, which sufficiently exploits the gradient of the Q function yet tends to collapse into a unimodal policy with low diversity. To address this issue, we propose CGPO, \textbf{C}ritic-\textbf{G}uided diffusion \textbf{P}olicy \textbf{O}ptimization, which effectively balances exploration and exploitation with the training-free guidance technique integrated into the denoising process of diffusion policy. Concretely, CGPO steers action generation toward high-value regions defined by the critic network and uses the guided actions as regression objectives. In this manner, CGPO reduces the time required to obtain high-quality actions and improves final performance with better balance between the exploration-exploitation tradeoff. We validate the effectiveness of CGPO on 5 MuJoCo locomotion tasks, and CGPO achieves state-of-the-art performance compared with existing diffusion-based RL methods. Notably, CGPO is the first success to incorporate diffusion policy into real-world RL, with its superior performance on Franka robot arm grasping tasks. Our official page is released at https://dingsht.tech/cgpo-webpage.

Chaofan Huo, Ye Shi, Yuexin Ma et al.

Modeling and capturing the 3D spatial arrangement of the human and the object is the key to perceiving 3D human-object interaction from monocular images. In this work, we propose to use the Human-Object Offset between anchors which are densely sampled from the surface of human mesh and object mesh to represent human-object spatial relation. Compared with previous works which use contact map or implicit distance filed to encode 3D human-object spatial relations, our method is a simple and efficient way to encode the highly detailed spatial correlation between the human and object. Based on this representation, we propose Stacked Normalizing Flow (StackFLOW) to infer the posterior distribution of human-object spatial relations from the image. During the optimization stage, we finetune the human body pose and object 6D pose by maximizing the likelihood of samples based on this posterior distribution and minimizing the 2D-3D corresponding reprojection loss. Extensive experimental results show that our method achieves impressive results on two challenging benchmarks, BEHAVE and InterCap datasets.

Kaiyang Ji, Bingsheng Qian, Binghuan Wu et al.

We study real-time audio-responsive character control as a deployment-faithful problem: strictly causal, bounded-latency streaming that must generate coherent full-body motion at interactive frame rates while the audio condition can change abruptly, including tempo shifts, drops, or user edits. Prior music-to-motion systems are largely optimized for offline generation with global context, and degrade in streaming rollouts where conditioning history becomes stale or unreliable. We introduce DiscoForcing, a streaming audio-driven diffusion framework that combines a causal music encoder that captures rhythmic structure and phase dynamics with a diffusion-forcing sequence model trained under heterogeneous noise levels across the temporal horizon. Building on this, we design a hybrid temporal schedule and a history-guided streaming sampler to explicitly trade off responsiveness against long-horizon consistency under non-stationary audio. Implemented in an end-to-end real-time interactive system with online avatar playback and humanoid deployment workflows, DiscoForcing delivers more stable long-horizon rollouts and sharper audio-motion alignment than prior baselines under matched causality and latency constraints while maintaining real-time throughput.

Lin Wang, Zhichao Wang, Ye Shi et al.

Federated Learning (FL) enables collaborative model training across distributed devices while preserving data privacy. Nonetheless, the heterogeneity of edge devices often leads to inconsistent performance of the globally trained models, resulting in unfair outcomes among users. Existing federated fairness algorithms strive to enhance fairness but often fall short in maintaining the overall performance of the global model, typically measured by the average accuracy across all clients. To address this issue, we propose a novel algorithm that leverages entropy-based aggregation combined with model and gradient alignments to simultaneously optimize fairness and global model performance. Our method employs a bi-level optimization framework, where we derive an analytic solution to the aggregation probability in the inner loop, making the optimization process computationally efficient. Additionally, we introduce an innovative alignment update and an adaptive strategy in the outer loop to further balance global model's performance and fairness. Theoretical analysis indicates that our approach guarantees convergence even in non-convex FL settings and demonstrates significant fairness improvements in generalized regression and strongly convex models. Empirically, our approach surpasses state-of-the-art federated fairness algorithms, ensuring consistent performance among clients while improving the overall performance of the global model.

Leijie Zhang, Ye Shi, Yu-Cheng Chang et al.

Distributed fuzzy neural networks (DFNNs) have attracted increasing attention recently due to their learning abilities in handling data uncertainties in distributed scenarios. However, it is challenging for DFNNs to handle cases in which the local data are non-independent and identically distributed (non-IID). In this paper, we propose a federated fuzzy neural network (FedFNN) with evolutionary rule learning (ERL) to cope with non-IID issues as well as data uncertainties. The FedFNN maintains a global set of rules in a server and a personalized subset of these rules for each local client. ERL is inspired by the theory of biological evolution; it encourages rule variations while activating superior rules and deactivating inferior rules for local clients with non-IID data. Specifically, ERL consists of two stages in an iterative procedure: a rule cooperation stage that updates global rules by aggregating local rules based on their activation statuses and a rule evolution stage that evolves the global rules and updates the activation statuses of the local rules. This procedure improves both the generalization and personalization of the FedFNN for dealing with non-IID issues and data uncertainties. Extensive experiments conducted on a range of datasets demonstrate the superiority of the FedFNN over state-of-the-art methods.

Lijun Sun, Yu-Cheng Chang, Chao Lyu et al.

The multiple-target self-organizing pursuit (SOP) problem has wide applications and has been considered a challenging self-organization game for distributed systems, in which intelligent agents cooperatively pursue multiple dynamic targets with partial observations. This work proposes a framework for decentralized multi-agent systems to improve the implicit coordination capabilities in search and pursuit. We model a self-organizing system as a partially observable Markov game (POMG) featured by large-scale, decentralization, partial observation, and noncommunication. The proposed distributed algorithm: fuzzy self-organizing cooperative coevolution (FSC2) is then leveraged to resolve the three challenges in multi-target SOP: distributed self-organizing search (SOS), distributed task allocation, and distributed single-target pursuit. FSC2 includes a coordinated multi-agent deep reinforcement learning (MARL) method that enables homogeneous agents to learn natural SOS patterns. Additionally, we propose a fuzzy-based distributed task allocation method, which locally decomposes multi-target SOP into several single-target pursuit problems. The cooperative coevolution principle is employed to coordinate distributed pursuers for each single-target pursuit problem. Therefore, the uncertainties of inherent partial observation and distributed decision-making in the POMG can be alleviated. The experimental results demonstrate that by decomposing the SOP task, FSC2 achieves superior performance compared with other implicit coordination policies fully trained by general MARL algorithms. The scalability of FSC2 is proved that up to 2048 FSC2 agents perform efficient multi-target SOP with almost 100 percent capture rates. Empirical analyses and ablation studies verify the interpretability, rationality, and effectiveness of component algorithms in FSC2.

Shutong Ding, Jingya Wang, Yali Du et al.

Recent advances in constrained reinforcement learning (RL) have endowed reinforcement learning with certain safety guarantees. However, deploying existing constrained RL algorithms in continuous control tasks with general hard constraints remains challenging, particularly in those situations with non-convex hard constraints. Inspired by the generalized reduced gradient (GRG) algorithm, a classical constrained optimization technique, we propose a reduced policy optimization (RPO) algorithm that combines RL with GRG to address general hard constraints. RPO partitions actions into basic actions and nonbasic actions following the GRG method and outputs the basic actions via a policy network. Subsequently, RPO calculates the nonbasic actions by solving equations based on equality constraints using the obtained basic actions. The policy network is then updated by implicitly differentiating nonbasic actions with respect to basic actions. Additionally, we introduce an action projection procedure based on the reduced gradient and apply a modified Lagrangian relaxation technique to ensure inequality constraints are satisfied. To the best of our knowledge, RPO is the first attempt that introduces GRG to RL as a way of efficiently handling both equality and inequality hard constraints. It is worth noting that there is currently a lack of RL environments with complex hard constraints, which motivates us to develop three new benchmarks: two robotics manipulation tasks and a smart grid operation control task. With these benchmarks, RPO achieves better performance than previous constrained RL algorithms in terms of both cumulative reward and constraint violation. We believe RPO, along with the new benchmarks, will open up new opportunities for applying RL to real-world problems with complex constraints.

Chaofan Huo, Ye Shi, Jingya Wang

Learning the prior knowledge of the 3D human-object spatial relation is crucial for reconstructing human-object interaction from images and understanding how humans interact with objects in 3D space. Previous works learn this prior from datasets collected in controlled environments, but due to the diversity of domains, they struggle to generalize to real-world scenarios. To overcome this limitation, we present a 2D-supervised method that learns the 3D human-object spatial relation prior purely from 2D images in the wild. Our method utilizes a flow-based neural network to learn the prior distribution of the 2D human-object keypoint layout and viewports for each image in the dataset. The effectiveness of the prior learned from 2D images is demonstrated on the human-object reconstruction task by applying the prior to tune the relative pose between the human and the object during the post-optimization stage. To validate and benchmark our method on in-the-wild images, we collect the WildHOI dataset from the YouTube website, which consists of various interactions with 8 objects in real-world scenarios. We conduct the experiments on the indoor BEHAVE dataset and the outdoor WildHOI dataset. The results show that our method achieves almost comparable performance with fully 3D supervised methods on the BEHAVE dataset, even if we have only utilized the 2D layout information, and outperforms previous methods in terms of generality and interaction diversity on in-the-wild images.

Yilan Dong, Chunlin Yu, Ruiyang Ha et al.

Existing gait recognition benchmarks mostly include minor clothing variations in the laboratory environments, but lack persistent changes in appearance over time and space. In this paper, we propose the first in-the-wild benchmark CCGait for cloth-changing gait recognition, which incorporates diverse clothing changes, indoor and outdoor scenes, and multi-modal statistics over 92 days. To further address the coupling effect of clothing and viewpoint variations, we propose a hybrid approach HybridGait that exploits both temporal dynamics and the projected 2D information of 3D human meshes. Specifically, we introduce a Canonical Alignment Spatial-Temporal Transformer (CA-STT) module to encode human joint position-aware features, and fully exploit 3D dense priors via a Silhouette-guided Deformation with 3D-2D Appearance Projection (SilD) strategy. Our contributions are twofold: we provide a challenging benchmark CCGait that captures realistic appearance changes across an expanded and space, and we propose a hybrid framework HybridGait that outperforms prior works on CCGait and Gait3D benchmarks. Our project page is available at https://github.com/HCVLab/HybridGait.

Zeyuan Wang, Da Li, Yulin Chen et al.

Online off-policy reinforcement learning (RL) is shaped by two coupled choices: the policy class and the update rule. Gaussian policies are fast and have tractable entropy, but struggle with multimodal action distributions. Generative policies are more expressive, but often require iterative sampling or lack tractable entropy estimates. On the optimisation side, SAC-style soft policy improvement and mirror descent (MD) can be viewed as minimising different KL divergences: the former moves the policy towards a value-induced Boltzmann distribution, while the latter regularises each update against the previous policy. Combining entropy regularisation with an MD constraint is therefore attractive, as it supports exploration while stabilising policy improvement; however, the resulting target can be multimodal and is poorly matched by unimodal Gaussian policies. We propose Stochastic MeanFlow Policies (SMFP), a one-step generative policy class that maps Gaussian noise to actions through a MeanFlow transformation. This stochastic reparameterisation yields a tractable entropy surrogate and allows MeanFlow policies to be trained within off-policy mirror descent under a unified objective for exploratory yet stable improvement. Across seven MuJoCo benchmarks, SMFP improves over Gaussian and generative baselines while retaining single-step inference efficiency.

Wanxing Chang, Ye Shi, Jingya Wang

Learning with noisy labels (LNL) poses a significant challenge in training a well-generalized model while avoiding overfitting to corrupted labels. Recent advances have achieved impressive performance by identifying clean labels and correcting corrupted labels for training. However, the current approaches rely heavily on the model's predictions and evaluate each sample independently without considering either the global and local structure of the sample distribution. These limitations typically result in a suboptimal solution for the identification and correction processes, which eventually leads to models overfitting to incorrect labels. In this paper, we propose a novel optimal transport (OT) formulation, called Curriculum and Structure-aware Optimal Transport (CSOT). CSOT concurrently considers the inter- and intra-distribution structure of the samples to construct a robust denoising and relabeling allocator. During the training process, the allocator incrementally assigns reliable labels to a fraction of the samples with the highest confidence. These labels have both global discriminability and local coherence. Notably, CSOT is a new OT formulation with a nonconvex objective function and curriculum constraints, so it is not directly compatible with classical OT solvers. Here, we develop a lightspeed computational method that involves a scaling iteration within a generalized conditional gradient framework to solve CSOT efficiently. Extensive experiments demonstrate the superiority of our method over the current state-of-the-arts in LNL. Code is available at https://github.com/changwxx/CSOT-for-LNL.

Shutong Ding, Yimiao Zhou, Ke Hu et al.

Recent advances in diffusion-based reinforcement learning (RL) methods have demonstrated promising results in a wide range of continuous control tasks. However, existing works in this field focus on the application of diffusion policies while leaving the diffusion critics unexplored. In fact, since policy optimization fundamentally relies on the critic, accurate value estimation is far more important than policy expressiveness. Furthermore, given the stochasticity of most reinforcement learning tasks, it has been confirmed that the critic is more appropriately depicted with a distributional model. Motivated by these points, we propose a novel distributional RL method with Diffusion Bridge Critics (DBC). DBC directly models the inverse cumulative distribution function (CDF) of the Q value. This allows us to accurately capture the value distribution and prevents it from collapsing into a trivial Gaussian distribution owing to the strong distribution-matching capability of the diffusion bridge. Moreover, we further derive an analytic integral formula to address discretization errors in DBC, which is essential in value estimation. To our knowledge, DBC is the first work to employ the diffusion bridge model as the critic. Notably, DBC is also a plug-and-play component and can be integrated into most existing RL frameworks. Experimental results on MuJoCo robot control benchmarks demonstrate the superiority of DBC compared with previous distributional critic models.

Zhaoyang Liu, Mokai Pan, Zhongyi Wang et al.

Imitation learning with diffusion models has advanced robotic control by capturing multi-modal action distributions. However, existing approaches typically treat observations as high-level conditioning inputs to the denoising network, rather than integrating them into the stochastic dynamics of the diffusion process itself. As a result, sampling must begin from random Gaussian noise, weakening the coupling between perception and control and often yielding suboptimal performance. We introduce BridgePolicy, a generative visuomotor policy that explicitly embeds observations within the stochastic differential equation via a diffusion-bridge formulation. By constructing an observation-informed trajectory, BridgePolicy enables sampling to start from a rich, informative prior rather than random noise, substantially improving precision and reliability in control. A key challenge is that classical diffusion bridges connect distributions with matched dimensionality, whereas robotic observations are heterogeneous and multi-modal and do not naturally align with the action space. To address this, we design a multi-modal fusion module and a semantic aligner that unify visual and state inputs and align observation and action representations, making the bridge applicable to heterogeneous robot data. Extensive experiments across 52 simulation tasks on three benchmarks and five real-world tasks demonstrate that BridgePolicy consistently outperforms state-of-the-art generative policies.

Bikang Pan, Wei Huang, Ye Shi

Integrating pretrained vision-language foundation models like CLIP into federated learning has attracted significant attention for enhancing generalization across diverse tasks. Typically, federated learning of vision-language models employs prompt learning to reduce communication and computational costs, i.e., prompt-based federated learning. However, there is limited theoretical analysis to understand the performance of prompt-based federated learning. In this work, we construct a theoretical analysis framework for prompt-based federated learning via feature learning theory. Specifically, we monitor the evolution of signal learning and noise memorization in prompt-based federated learning, demonstrating that performance can be assessed by the ratio of task-relevant to task-irrelevant coefficients. Furthermore, we draw an analogy between income and risk in portfolio optimization and the task-relevant and task-irrelevant terms in feature learning. Leveraging inspiration from portfolio optimization that combining two independent assets will maintain the income while reducing the risk, we introduce two prompts: global prompt and local prompt to construct a prompt portfolio to balance the generalization and personalization. Consequently, we showed the performance advantage of the prompt portfolio and derived the optimal mixing coefficient. These theoretical claims have been further supported by empirical experiments.

Jiebao Zhang, Haoyu Yan, Haoyu Wang et al.

Learning to solve the Alternating Current Optimal Power Flow (AC-OPF) problem by neural networks (NNs) is a promising approach in real-time applications. Existing methods to ensure the physical feasibility of NN outputs embed a power flow (PF) solver within networks. However, the gradient through the PF solver, namely, implicit differentiation, needs manual Jacobian derivation and the solution of linear systems, which is computationally prohibitive and hinders integration with modern automatic differentiation (AD) frameworks. To address these challenges, we propose FPL-OPF, a novel unsupervised learning framework that incorporates a Fast Physics-aware Layer for AC-OPF problems. FPL-OPF embeds a fast PF iterative solver within the NN and takes solely the last few or even the final iterations into the AD graph. This design ensures high computational efficiency for both the forward and backward passes, circumventing complex custom backward implementations. Theoretically, we rigorously prove that the gradient from this design serves as a high-fidelity surrogate of the true implicit gradient under mild conditions. Extensive experiments demonstrate that FPL-OPF achieves significant speedups over state-of-the-art unsupervised learning approaches, while maintaining near-zero constraint violations and competitive optimality. Our code is available at https://github.com/wowotou1998/fpl-opf

Kaizhen Zhu, Mokai Pan, Yuexin Ma et al.

Recent advances in diffusion bridge models leverage Doob's $h$-transform to establish fixed endpoints between distributions, demonstrating promising results in image translation and restoration tasks. However, these approaches frequently produce blurred or excessively smoothed image details and lack a comprehensive theoretical foundation to explain these shortcomings. To address these limitations, we propose UniDB, a unified framework for diffusion bridges based on Stochastic Optimal Control (SOC). UniDB formulates the problem through an SOC-based optimization and derives a closed-form solution for the optimal controller, thereby unifying and generalizing existing diffusion bridge models. We demonstrate that existing diffusion bridges employing Doob's $h$-transform constitute a special case of our framework, emerging when the terminal penalty coefficient in the SOC cost function tends to infinity. By incorporating a tunable terminal penalty coefficient, UniDB achieves an optimal balance between control costs and terminal penalties, substantially improving detail preservation and output quality. Notably, UniDB seamlessly integrates with existing diffusion bridge models, requiring only minimal code modifications. Extensive experiments across diverse image restoration tasks validate the superiority and adaptability of the proposed framework. Our code is available at https://github.com/UniDB-SOC/UniDB/.

Kaizhen Zhu, Mokai Pan, Zhechuan Yu et al.

Diffusion Bridge and Flow Matching have both demonstrated compelling empirical performance in transformation between arbitrary distributions. However, there remains confusion about which approach is generally preferable, and the substantial discrepancies in their modeling assumptions and practical implementations have hindered a unified theoretical account of their relative merits. We have, for the first time, provided a unified theoretical and experimental validation of these two models. We recast their frameworks through the lens of Stochastic Optimal Control and prove that the cost function of the Diffusion Bridge is lower, guiding the system toward more stable and natural trajectories. Simultaneously, from the perspective of Optimal Transport, interpolation coefficients $t$ and $1-t$ of Flow Matching become increasingly ineffective when the training data size is reduced. To corroborate these theoretical claims, we propose a novel, powerful architecture for Diffusion Bridge built on a latent Transformer, and implement a Flow Matching model with the same structure to enable a fair performance comparison in various experiments. Comprehensive experiments are conducted across Image Inpainting, Super-Resolution, Deblurring, Denoising, Translation, and Style Transfer tasks, systematically varying both the distributional discrepancy (different difficulty) and the training data size. Extensive empirical results align perfectly with our theoretical predictions and allow us to delineate the respective advantages and disadvantages of these two models. Our code is available at https://anonymous.4open.science/r/DBFM-3E8E/.

Fan Liu, Bikang Pan, Zhongyi Wang et al.

Federated instruction tuning of large language models (LLMs) is challenged by significant data heterogeneity across clients, demanding robust personalization. The Mixture of Experts (MoE) architecture, where experts can specialize in distinct data patterns, presents a natural architectural solution to this challenge. The inherent sparsity of the MoE architecture, achieved by selectively activating experts, poses a significant challenge to its integration with federated learning (FL). Conventional FL frameworks, designed for dense models, naively aggregate all expert parameters irrespective of their local activation patterns. This naive approach not only undermines MoE's dynamic sparsity but also risks corrupting the world knowledge within pretrained experts. To address this, we propose FLEx (Federated LLMs with Personalized Experts), a novel framework that leverages pretrained MoE-based LLMs for efficient personalization. By aggregating only the shared non-expert parameters, FLEx significantly reduces communication overhead and preserves the world knowledge stored within the frozen pretrained experts. For personalization, we introduce a novel expert grafting mechanism that leverages dynamic sparsity to construct a client-specific expert from selected components of pretrained experts, tailored to local data. This grafted expert is then fine-tuned locally alongside the gating mechanism. This joint training enables the model to learn when to leverage the shared knowledge from frozen experts and when to employ the personalized one. Evaluations on diverse, non-IID instruction tuning datasets show that FLEx consistently outperforms federated baselines on average, while demonstrating strong knowledge preservation on the knowledge-driven benchmark MMLU. Our code is available at \href{https://anonymous.4open.science/r/FLEx-8F12}{\texttt{https://anonymous.4open.science/r/FLEx-8F12}}.

Mokai Pan, Kaizhen Zhu, Yuexin Ma et al.

Recent advances in diffusion bridge models leverage Doob's $h$-transform to establish fixed endpoints between distributions, demonstrating promising results in image translation and restoration tasks. However, these approaches often produce blurred or excessively smoothed image details and lack a comprehensive theoretical foundation to explain these shortcomings. To address these limitations, we propose UniDB, a unified and fast-sampling framework for diffusion bridges based on Stochastic Optimal Control (SOC). We reformulate the problem through an SOC-based optimization, proving that existing diffusion bridges employing Doob's $h$-transform constitute a special case, emerging when the terminal penalty coefficient in the SOC cost function tends to infinity. By incorporating a tunable terminal penalty coefficient, UniDB achieves an optimal balance between control costs and terminal penalties, substantially improving detail preservation and output quality. To avoid computationally expensive costs of iterative Euler sampling methods in UniDB, we design a training-free accelerated algorithm by deriving exact closed-form solutions for UniDB's reverse-time SDE. It is further complemented by replacing conventional noise prediction with a more stable data prediction model, along with an SDE-Corrector mechanism that maintains perceptual quality for low-step regimes, effectively reducing error accumulation. Extensive experiments across diverse image restoration tasks validate the superiority and adaptability of the proposed framework, bridging the gap between theoretical generality and practical efficiency. Our code is available online https://github.com/2769433owo/UniDB-plusplus.

Zeyuan Wang, Da Li, Yulin Chen et al.

We introduce a one-step generative policy for offline reinforcement learning that maps noise directly to actions via a residual reformulation of MeanFlow, making it compatible with Q-learning. While one-step Gaussian policies enable fast inference, they struggle to capture complex, multimodal action distributions. Existing flow-based methods improve expressivity but typically rely on distillation and two-stage training when trained with Q-learning. To overcome these limitations, we propose to reformulate MeanFlow to enable direct noise-to-action generation by integrating the velocity field and noise-to-action transformation into a single policy network-eliminating the need for separate velocity estimation. We explore several reformulation variants and identify an effective residual formulation that supports expressive and stable policy learning. Our method offers three key advantages: 1) efficient one-step noise-to-action generation, 2) expressive modelling of multimodal action distributions, and 3) efficient and stable policy learning via Q-learning in a single-stage training setup. Extensive experiments on 73 tasks across the OGBench and D4RL benchmarks demonstrate that our method achieves strong performance in both offline and offline-to-online reinforcement learning settings. Code is available at https://github.com/HiccupRL/MeanFlowQL.

Yiming Zhong, Yumeng Liu, Chuyang Xiao et al.

Learning effective visuomotor policies for robotic manipulation is challenging, as it requires generating precise actions while maintaining computational efficiency. Existing methods remain unsatisfactory due to inherent limitations in the essential action representation and the basic network architectures. We observe that representing actions in the frequency domain captures the structured nature of motion more effectively: low-frequency components reflect global movement patterns, while high-frequency components encode fine local details. Additionally, robotic manipulation tasks of varying complexity demand different levels of modeling precision across these frequency bands. Motivated by this, we propose a novel paradigm for visuomotor policy learning that progressively models hierarchical frequency components. To further enhance precision, we introduce continuous latent representations that maintain smoothness and continuity in the action space. Extensive experiments across diverse 2D and 3D robotic manipulation benchmarks demonstrate that our approach outperforms existing methods in both accuracy and efficiency, showcasing the potential of a frequency-domain autoregressive framework with continuous tokens for generalized robotic manipulation.Code is available at https://github.com/4DVLab/Freqpolicy

Wonhyeok Choi, Shutong Ding, Minwoo Choi et al.

Diffusion policies have emerged as a powerful approach for robotic control, demonstrating superior expressiveness in modeling multimodal action distributions compared to conventional policy networks. However, their integration with online reinforcement learning remains challenging due to fundamental incompatibilities between diffusion model training objectives and standard RL policy improvement mechanisms. This paper presents the first comprehensive review and empirical analysis of current Online Diffusion Policy Reinforcement Learning (Online DPRL) algorithms for scalable robotic control systems. We propose a novel taxonomy that categorizes existing approaches into four distinct families--Action-Gradient, Q-Weighting, Proximity-Based, and Backpropagation Through Time (BPTT) methods--based on their policy improvement mechanisms. Through extensive experiments on a unified NVIDIA Isaac Lab benchmark encompassing 12 diverse robotic tasks, we systematically evaluate representative algorithms across five critical dimensions: task diversity, parallelization capability, diffusion step scalability, cross-embodiment generalization, and environmental robustness. Our analysis identifies key findings regarding the fundamental trade-offs inherent in each algorithmic family, particularly concerning sample efficiency and scalability. Furthermore, we reveal critical computational and algorithmic bottlenecks that currently limit the practical deployment of online DPRL. Based on these findings, we provide concrete guidelines for algorithm selection tailored to specific operational constraints and outline promising future research directions to advance the field toward more general and scalable robotic learning systems.

Lingxiao Yang, Shutong Ding, Yifan Cai et al.

Recent advances in diffusion models attempt to handle conditional generative tasks by utilizing a differentiable loss function for guidance without the need for additional training. While these methods achieved certain success, they often compromise on sample quality and require small guidance step sizes, leading to longer sampling processes. This paper reveals that the fundamental issue lies in the manifold deviation during the sampling process when loss guidance is employed. We theoretically show the existence of manifold deviation by establishing a certain lower bound for the estimation error of the loss guidance. To mitigate this problem, we propose Diffusion with Spherical Gaussian constraint (DSG), drawing inspiration from the concentration phenomenon in high-dimensional Gaussian distributions. DSG effectively constrains the guidance step within the intermediate data manifold through optimization and enables the use of larger guidance steps. Furthermore, we present a closed-form solution for DSG denoising with the Spherical Gaussian constraint. Notably, DSG can seamlessly integrate as a plugin module within existing training-free conditional diffusion methods. Implementing DSG merely involves a few lines of additional code with almost no extra computational overhead, yet it leads to significant performance improvements. Comprehensive experimental results in various conditional generation tasks validate the superiority and adaptability of DSG in terms of both sample quality and time efficiency.

Ye Shi, C. S. George Lee

Irregular distribution in latent space causes posterior collapse, misalignment between posterior and prior, and ill-sampling problem in Variational Autoencoders (VAEs). In this paper, we introduce a novel adaptable three-stage Uniform Transformation (UT) module -- Gaussian Kernel Density Estimation (G-KDE) clustering, non-parametric Gaussian Mixture (GM) Modeling, and Probability Integral Transform (PIT) -- to address irregular latent distributions. By reconfiguring irregular distributions into a uniform distribution in the latent space, our approach significantly enhances the disentanglement and interpretability of latent representations, overcoming the limitation of traditional VAE models in capturing complex data structures. Empirical evaluations demonstrated the efficacy of our proposed UT module in improving disentanglement metrics across benchmark datasets -- dSprites and MNIST. Our findings suggest a promising direction for advancing representation learning techniques, with implication for future research in extending this framework to more sophisticated datasets and downstream tasks.

Qianyang Wu, Ye Shi, Xiaoshui Huang et al.

This paper addresses new methodologies to deal with the challenging task of generating dynamic Human-Object Interactions from textual descriptions (Text2HOI). While most existing works assume interactions with limited body parts or static objects, our task involves addressing the variation in human motion, the diversity of object shapes, and the semantic vagueness of object motion simultaneously. To tackle this, we propose a novel Text-guided Human-Object Interaction diffusion model with Relation Intervention (THOR). THOR is a cohesive diffusion model equipped with a relation intervention mechanism. In each diffusion step, we initiate text-guided human and object motion and then leverage human-object relations to intervene in object motion. This intervention enhances the spatial-temporal relations between humans and objects, with human-centric interaction representation providing additional guidance for synthesizing consistent motion from text. To achieve more reasonable and realistic results, interaction losses is introduced at different levels of motion granularity. Moreover, we construct Text-BEHAVE, a Text2HOI dataset that seamlessly integrates textual descriptions with the currently largest publicly available 3D HOI dataset. Both quantitative and qualitative experiments demonstrate the effectiveness of our proposed model.

Hongxia Li, Wei Huang, Jingya Wang et al.

Prompt learning in pretrained visual-language models has shown remarkable flexibility across various downstream tasks. Leveraging its inherent lightweight nature, recent research attempted to integrate the powerful pretrained models into federated learning frameworks to simultaneously reduce communication costs and promote local training on insufficient data. Despite these efforts, current federated prompt learning methods lack specialized designs to systematically address severe data heterogeneities, e.g., data distribution with both label and feature shifts involved. To address this challenge, we present Federated Prompts Cooperation via Optimal Transport (FedOTP), which introduces efficient collaborative prompt learning strategies to capture diverse category traits on a per-client basis. Specifically, for each client, we learn a global prompt to extract consensus knowledge among clients, and a local prompt to capture client-specific category characteristics. Unbalanced Optimal Transport is then employed to align local visual features with these prompts, striking a balance between global consensus and local personalization. By relaxing one of the equality constraints, FedOTP enables prompts to focus solely on the core regions of image patches. Extensive experiments on datasets with various types of heterogeneities have demonstrated that our FedOTP outperforms the state-of-the-art methods.

Juze Zhang, Jingyan Zhang, Zining Song et al.

Humans naturally interact with both others and the surrounding multiple objects, engaging in various social activities. However, recent advances in modeling human-object interactions mostly focus on perceiving isolated individuals and objects, due to fundamental data scarcity. In this paper, we introduce HOI-M3, a novel large-scale dataset for modeling the interactions of Multiple huMans and Multiple objects. Notably, it provides accurate 3D tracking for both humans and objects from dense RGB and object-mounted IMU inputs, covering 199 sequences and 181M frames of diverse humans and objects under rich activities. With the unique HOI-M3 dataset, we introduce two novel data-driven tasks with companion strong baselines: monocular capture and unstructured generation of multiple human-object interactions. Extensive experiments demonstrate that our dataset is challenging and worthy of further research about multiple human-object interactions and behavior analysis. Our HOI-M3 dataset, corresponding codes, and pre-trained models will be disseminated to the community for future research.

Tianyu Cui, Hongxia Li, Jingya Wang et al.

Federated Prompt Learning (FPL) incorporates large pre-trained Vision-Language models (VLM) into federated learning through prompt tuning. The transferable representations and remarkable generalization capacity of VLM make them highly compatible with the integration of federated learning. Addressing data heterogeneity in federated learning requires personalization, but excessive focus on it across clients could compromise the model's ability to generalize effectively. To preserve the impressive generalization capability of VLM, it is crucial to strike a balance between personalization and generalization in FPL. To tackle this challenge, we proposed Federated Prompt Learning with CLIP Generalization and low-rank Personalization (FedPGP), which employs pre-trained CLIP to provide knowledge-guidance on the global prompt for improved generalization and incorporates a low-rank adaptation term to personalize the global prompt. Further, FedPGP integrates a prompt-wise contrastive loss to achieve knowledge guidance and personalized adaptation simultaneously, enabling a harmonious balance between personalization and generalization in FPL. We conduct extensive experiments on various datasets to explore base-to-novel generalization in both category-level and domain-level scenarios with heterogeneous data, showing the superiority of FedPGP in balancing generalization and personalization.

Jiangnan Tang, Jingya Wang, Kaiyang Ji et al.

Estimating full-body human motion via sparse tracking signals from head-mounted displays and hand controllers in 3D scenes is crucial to applications in AR/VR. One of the biggest challenges to this task is the one-to-many mapping from sparse observations to dense full-body motions, which endowed inherent ambiguities. To help resolve this ambiguous problem, we introduce a new framework to combine rich contextual information provided by scenes to benefit full-body motion tracking from sparse observations. To estimate plausible human motions given sparse tracking signals and 3D scenes, we develop $\text{S}^2$Fusion, a unified framework fusing \underline{S}cene and sparse \underline{S}ignals with a conditional dif\underline{Fusion} model. $\text{S}^2$Fusion first extracts the spatial-temporal relations residing in the sparse signals via a periodic autoencoder, and then produces time-alignment feature embedding as additional inputs. Subsequently, by drawing initial noisy motion from a pre-trained prior, $\text{S}^2$Fusion utilizes conditional diffusion to fuse scene geometry and sparse tracking signals to generate full-body scene-aware motions. The sampling procedure of $\text{S}^2$Fusion is further guided by a specially designed scene-penetration loss and phase-matching loss, which effectively regularizes the motion of the lower body even in the absence of any tracking signals, making the generated motion much more plausible and coherent. Extensive experimental results have demonstrated that our $\text{S}^2$Fusion outperforms the state-of-the-art in terms of estimation quality and smoothness.

Bin Li, Ye Shi, Qian Yu et al.

Unsupervised cross-domain image retrieval (UCIR) aims to retrieve images sharing the same category across diverse domains without relying on labeled data. Prior approaches have typically decomposed the UCIR problem into two distinct tasks: intra-domain representation learning and cross-domain feature alignment. However, these segregated strategies overlook the potential synergies between these tasks. This paper introduces ProtoOT, a novel Optimal Transport formulation explicitly tailored for UCIR, which integrates intra-domain feature representation learning and cross-domain alignment into a unified framework. ProtoOT leverages the strengths of the K-means clustering method to effectively manage distribution imbalances inherent in UCIR. By utilizing K-means for generating initial prototypes and approximating class marginal distributions, we modify the constraints in Optimal Transport accordingly, significantly enhancing its performance in UCIR scenarios. Furthermore, we incorporate contrastive learning into the ProtoOT framework to further improve representation learning. This encourages local semantic consistency among features with similar semantics, while also explicitly enforcing separation between features and unmatched prototypes, thereby enhancing global discriminativeness. ProtoOT surpasses existing state-of-the-art methods by a notable margin across benchmark datasets. Notably, on DomainNet, ProtoOT achieves an average P@200 enhancement of 18.17%, and on Office-Home, it demonstrates a P@15 improvement of 3.83%.

Chunlin Yu, Ye Shi, Jingya Wang

Previous endeavors in self-supervised learning have enlightened the research of deep clustering from an instance discrimination perspective. Built upon this foundation, recent studies further highlight the importance of grouping semantically similar instances. One effective method to achieve this is by promoting the semantic structure preserved by neighborhood consistency. However, the samples in the local neighborhood may be limited due to their close proximity to each other, which may not provide substantial and diverse supervision signals. Inspired by the versatile re-ranking methods in the context of image retrieval, we propose to employ an efficient online re-ranking process to mine more informative neighbors in a Contextually Affinitive (ConAff) Neighborhood, and then encourage the cross-view neighborhood consistency. To further mitigate the intrinsic neighborhood noises near cluster boundaries, we propose a progressively relaxed boundary filtering strategy to circumvent the issues brought by noisy neighbors. Our method can be easily integrated into the generic self-supervised frameworks and outperforms the state-of-the-art methods on several popular benchmarks.

Zhongyi Cai, Ye Shi, Wei Huang et al.

Federated Learning (FL) has emerged as a promising distributed learning paradigm that enables multiple clients to learn a global model collaboratively without sharing their private data. However, the effectiveness of FL is highly dependent on the quality of the data that is being used for training. In particular, data heterogeneity issues, such as label distribution skew and feature skew, can significantly impact the performance of FL. Previous studies in FL have primarily focused on addressing label distribution skew data heterogeneity, while only a few recent works have made initial progress in tackling feature skew issues. Notably, these two forms of data heterogeneity have been studied separately and have not been well explored within a unified FL framework. To address this gap, we propose Fed-CO$_{2}$, a universal FL framework that handles both label distribution skew and feature skew within a \textbf{C}ooperation mechanism between the \textbf{O}nline and \textbf{O}ffline models. Specifically, the online model learns general knowledge that is shared among all clients, while the offline model is trained locally to learn the specialized knowledge of each individual client. To further enhance model cooperation in the presence of feature shifts, we design an intra-client knowledge transfer mechanism that reinforces mutual learning between the online and offline models, and an inter-client knowledge transfer mechanism to increase the models' domain generalization ability. Extensive experiments show that our Fed-CO$_{2}$ outperforms a wide range of existing personalized federated learning algorithms in terms of handling label distribution skew and feature skew, both individually and collectively. The empirical results are supported by our convergence analyses in a simplified setting.

Chunlin Yu, Hanqing Wang, Ye Shi et al.

3D affordance segmentation aims to link human instructions to touchable regions of 3D objects for embodied manipulations. Existing efforts typically adhere to single-object, single-affordance paradigms, where each affordance type or explicit instruction strictly corresponds to a specific affordance region and are unable to handle long-horizon tasks. Such a paradigm cannot actively reason about complex user intentions that often imply sequential affordances. In this paper, we introduce the Sequential 3D Affordance Reasoning task, which extends the traditional paradigm by reasoning from cumbersome user intentions and then decomposing them into a series of segmentation maps. Toward this, we construct the first instruction-based affordance segmentation benchmark that includes reasoning over both single and sequential affordances, comprising 180K instruction-point cloud pairs. Based on the benchmark, we propose our model, SeqAfford, to unlock the 3D multi-modal large language model with additional affordance segmentation abilities, which ensures reasoning with world knowledge and fine-grained affordance grounding in a cohesive framework. We further introduce a multi-granular language-point integration module to endow 3D dense prediction. Extensive experimental evaluations show that our model excels over well-established methods and exhibits open-world generalization with sequential reasoning abilities.

Zhenhao Zhang, Ye Shi, Lingxiao Yang et al.

Understanding and synthesizing realistic 3D hand-object interactions (HOI) is critical for applications ranging from immersive AR/VR to dexterous robotics. Existing methods struggle with generalization, performing well on closed-set objects and predefined tasks but failing to handle unseen objects or open-vocabulary instructions. We introduce OpenHOI, the first framework for open-world HOI synthesis, capable of generating long-horizon manipulation sequences for novel objects guided by free-form language commands. Our approach integrates a 3D Multimodal Large Language Model (MLLM) fine-tuned for joint affordance grounding and semantic task decomposition, enabling precise localization of interaction regions (e.g., handles, buttons) and breakdown of complex instructions (e.g., "Find a water bottle and take a sip") into executable sub-tasks. To synthesize physically plausible interactions, we propose an affordance-driven diffusion model paired with a training-free physics refinement stage that minimizes penetration and optimizes affordance alignment. Evaluations across diverse scenarios demonstrate OpenHOI's superiority over state-of-the-art methods in generalizing to novel object categories, multi-stage tasks, and complex language instructions. Our project page at \href{https://openhoi.github.io}

Bikang Pan, Qun Li, Xiaoying Tang et al.

The emergence of vision-language foundation models, such as CLIP, has revolutionized image-text representation, enabling a broad range of applications via prompt learning. Despite its promise, real-world datasets often contain noisy labels that can degrade prompt learning performance. In this paper, we demonstrate that using mean absolute error (MAE) loss in prompt learning, named PromptMAE, significantly enhances robustness against noisy labels while maintaining high accuracy. Though MAE is straightforward and recognized for its robustness, it is rarely used in noisy-label learning due to its slow convergence and poor performance outside prompt learning scenarios. To elucidate the robustness of PromptMAE, we leverage feature learning theory to show that MAE can suppress the influence of noisy samples, thereby improving the signal-to-noise ratio and enhancing overall robustness. Additionally, we introduce PromptOT, a prompt-based optimal transport data purification method to enhance the robustness further. PromptOT employs text features in vision-language models as prototypes to construct an optimal transportation matrix. This matrix effectively partitions datasets into clean and noisy subsets, allowing for the application of cross-entropy loss to the clean subset and MAE loss to the noisy subset. Our Noise-Label Prompt Learning method, named NLPrompt, offers a simple and efficient approach that leverages the expressive representations and precise alignment capabilities of vision-language models for robust prompt learning. We validate NLPrompt through extensive experiments across various noise settings, demonstrating significant performance improvements.

Yahao Fan, Tianxiang Gui, Kaiyang Ji et al.

Humanoid locomotion faces a critical scalability challenge: traditional reinforcement learning (RL) methods require task-specific rewards and struggle to leverage growing datasets, even as more training terrains are introduced. We propose DreamPolicy, a unified framework that enables a single policy to master diverse terrains and generalize zero-shot to unseen scenarios by systematically integrating offline data and diffusion-driven motion synthesis. At its core, DreamPolicy introduces Humanoid Motion Imagery (HMI) - future state predictions synthesized through an autoregressive terrain-aware diffusion planner curated by aggregating rollouts from specialized policies across various distinct terrains. Unlike human motion datasets requiring laborious retargeting, our data directly captures humanoid kinematics, enabling the diffusion planner to synthesize "dreamed" trajectories that encode terrain-specific physical constraints. These trajectories act as dynamic objectives for our HMI-conditioned policy, bypassing manual reward engineering and enabling cross-terrain generalization. DreamPolicy addresses the scalability limitations of prior methods: while traditional RL fails to exploit growing datasets, our framework scales seamlessly with more offline data. As the dataset expands, the diffusion prior learns richer locomotion skills, which the policy leverages to master new terrains without retraining. Experiments demonstrate that DreamPolicy achieves average 90% success rates in training environments and an average of 20% higher success on unseen terrains than the prevalent method. It also generalizes to perturbed and composite scenarios where prior approaches collapse. By unifying offline data, diffusion-based trajectory synthesis, and policy optimization, DreamPolicy overcomes the "one task, one policy" bottleneck, establishing a paradigm for scalable, data-driven humanoid control.

Tianyu Cui, Jinbin Bai, Guo-Hua Wang et al.

Evaluating image captions typically relies on reference captions, which are costly to obtain and exhibit significant diversity and subjectivity. While reference-free evaluation metrics have been proposed, most focus on cross-modal evaluation between captions and images. Recent research has revealed that the modality gap generally exists in the representation of contrastive learning-based multi-modal systems, undermining the reliability of cross-modality metrics like CLIPScore. In this paper, we propose CAMScore, a cyclic reference-free automatic evaluation metric for image captioning models. To circumvent the aforementioned modality gap, CAMScore utilizes a text-to-image model to generate images from captions and subsequently evaluates these generated images against the original images. Furthermore, to provide fine-grained information for a more comprehensive evaluation, we design a three-level evaluation framework for CAMScore that encompasses pixel-level, semantic-level, and objective-level perspectives. Extensive experiment results across multiple benchmark datasets show that CAMScore achieves a superior correlation with human judgments compared to existing reference-based and reference-free metrics, demonstrating the effectiveness of the framework.

Jie Tian, Ran Ji, Lingxiao Yang et al.

Gaze plays a crucial role in revealing human attention and intention, particularly in hand-object interaction scenarios, where it guides and synchronizes complex tasks that require precise coordination between the brain, hand, and object. Motivated by this, we introduce a novel task: Gaze-Guided Hand-Object Interaction Synthesis, with potential applications in augmented reality, virtual reality, and assistive technologies. To support this task, we present GazeHOI, the first dataset to capture simultaneous 3D modeling of gaze, hand, and object interactions. This task poses significant challenges due to the inherent sparsity and noise in gaze data, as well as the need for high consistency and physical plausibility in generating hand and object motions. To tackle these issues, we propose a stacked gaze-guided hand-object interaction diffusion model, named GHO-Diffusion. The stacked design effectively reduces the complexity of motion generation. We also introduce HOI-Manifold Guidance during the sampling stage of GHO-Diffusion, enabling fine-grained control over generated motions while maintaining the data manifold. Additionally, we propose a spatial-temporal gaze feature encoding for the diffusion condition and select diffusion results based on consistency scores between gaze-contact maps and gaze-interaction trajectories. Extensive experiments highlight the effectiveness of our method and the unique contributions of our dataset. More details in https://takiee.github.io/gaze-hoi/.