Fuchun Sun

Yan Zeng, Ruichu Cai, Fuchun Sun et al. · tsinghua

While Reinforcement Learning (RL) achieves tremendous success in sequential decision-making problems of many domains, it still faces key challenges of data inefficiency and the lack of interpretability. Interestingly, many researchers have leveraged insights from the causality literature recently, bringing forth flourishing works to unify the merits of causality and address well the challenges from RL. As such, it is of great necessity and significance to collate these Causal Reinforcement Learning (CRL) works, offer a review of CRL methods, and investigate the potential functionality from causality toward RL. In particular, we divide existing CRL approaches into two categories according to whether their causality-based information is given in advance or not. We further analyze each category in terms of the formalization of different models, ranging from the Markov Decision Process (MDP), Partially Observed Markov Decision Process (POMDP), Multi-Arm Bandits (MAB), and Dynamic Treatment Regime (DTR). Moreover, we summarize the evaluation matrices and open sources while we discuss emerging applications, along with promising prospects for the future development of CRL.

Yikai Wang, Xinghao Chen, Lele Cao et al.

Many adaptations of transformers have emerged to address the single-modal vision tasks, where self-attention modules are stacked to handle input sources like images. Intuitively, feeding multiple modalities of data to vision transformers could improve the performance, yet the inner-modal attentive weights may also be diluted, which could thus undermine the final performance. In this paper, we propose a multimodal token fusion method (TokenFusion), tailored for transformer-based vision tasks. To effectively fuse multiple modalities, TokenFusion dynamically detects uninformative tokens and substitutes these tokens with projected and aggregated inter-modal features. Residual positional alignment is also adopted to enable explicit utilization of the inter-modal alignments after fusion. The design of TokenFusion allows the transformer to learn correlations among multimodal features, while the single-modal transformer architecture remains largely intact. Extensive experiments are conducted on a variety of homogeneous and heterogeneous modalities and demonstrate that TokenFusion surpasses state-of-the-art methods in three typical vision tasks: multimodal image-to-image translation, RGB-depth semantic segmentation, and 3D object detection with point cloud and images. Our code is available at https://github.com/yikaiw/TokenFusion.

Ke Liang, Lingyuan Meng, Meng Liu et al.

Knowledge graph reasoning (KGR), aiming to deduce new facts from existing facts based on mined logic rules underlying knowledge graphs (KGs), has become a fast-growing research direction. It has been proven to significantly benefit the usage of KGs in many AI applications, such as question answering, recommendation systems, and etc. According to the graph types, existing KGR models can be roughly divided into three categories, i.e., static models, temporal models, and multi-modal models. Early works in this domain mainly focus on static KGR, and recent works try to leverage the temporal and multi-modal information, which are more practical and closer to real-world. However, no survey papers and open-source repositories comprehensively summarize and discuss models in this important direction. To fill the gap, we conduct a first survey for knowledge graph reasoning tracing from static to temporal and then to multi-modal KGs. Concretely, the models are reviewed based on bi-level taxonomy, i.e., top-level (graph types) and base-level (techniques and scenarios). Besides, the performances, as well as datasets, are summarized and presented. Moreover, we point out the challenges and potential opportunities to enlighten the readers. The corresponding open-source repository is shared on GitHub https://github.com/LIANGKE23/Awesome-Knowledge-Graph-Reasoning.

Chengliang Zhong, Peixing You, Xiaoxue Chen et al.

Detecting 3D keypoints from point clouds is important for shape reconstruction, while this work investigates the dual question: can shape reconstruction benefit 3D keypoint detection? Existing methods either seek salient features according to statistics of different orders or learn to predict keypoints that are invariant to transformation. Nevertheless, the idea of incorporating shape reconstruction into 3D keypoint detection is under-explored. We argue that this is restricted by former problem formulations. To this end, a novel unsupervised paradigm named SNAKE is proposed, which is short for shape-aware neural 3D keypoint field. Similar to recent coordinate-based radiance or distance field, our network takes 3D coordinates as inputs and predicts implicit shape indicators and keypoint saliency simultaneously, thus naturally entangling 3D keypoint detection and shape reconstruction. We achieve superior performance on various public benchmarks, including standalone object datasets ModelNet40, KeypointNet, SMPL meshes and scene-level datasets 3DMatch and Redwood. Intrinsic shape awareness brings several advantages as follows. (1) SNAKE generates 3D keypoints consistent with human semantic annotation, even without such supervision. (2) SNAKE outperforms counterparts in terms of repeatability, especially when the input point clouds are down-sampled. (3) the generated keypoints allow accurate geometric registration, notably in a zero-shot setting. Codes are available at https://github.com/zhongcl-thu/SNAKE

Xingzhe Su, Wenwen Qiang, Jie Hu et al.

Generative adversarial networks (GANs) have achieved remarkable progress in the natural image field. However, when applying GANs in the remote sensing (RS) image generation task, an extraordinary phenomenon is observed: the GAN model is more sensitive to the size of training data for RS image generation than for natural image generation. In other words, the generation quality of RS images will change significantly with the number of training categories or samples per category. In this paper, we first analyze this phenomenon from two kinds of toy experiments and conclude that the amount of feature information contained in the GAN model decreases with reduced training data. Then we establish a structural causal model (SCM) of the data generation process and interpret the generated data as the counterfactuals. Based on this SCM, we theoretically prove that the quality of generated images is positively correlated with the amount of feature information. This provides insights for enriching the feature information learned by the GAN model during training. Consequently, we propose two innovative adjustment schemes, namely Uniformity Regularization (UR) and Entropy Regularization (ER), to increase the information learned by the GAN model at the distributional and sample levels, respectively. We theoretically and empirically demonstrate the effectiveness and versatility of our methods. Extensive experiments on three RS datasets and two natural datasets show that our methods outperform the well-established models on RS image generation tasks. The source code is available at https://github.com/rootSue/Causal-RSGAN.

Runfa Chen, Yu Rong, Shangmin Guo et al.

After the great success of Vision Transformer variants (ViTs) in computer vision, it has also demonstrated great potential in domain adaptive semantic segmentation. Unfortunately, straightforwardly applying local ViTs in domain adaptive semantic segmentation does not bring in expected improvement. We find that the pitfall of local ViTs is due to the severe high-frequency components generated during both the pseudo-label construction and features alignment for target domains. These high-frequency components make the training of local ViTs very unsmooth and hurt their transferability. In this paper, we introduce a low-pass filtering mechanism, momentum network, to smooth the learning dynamics of target domain features and pseudo labels. Furthermore, we propose a dynamic of discrepancy measurement to align the distributions in the source and target domains via dynamic weights to evaluate the importance of the samples. After tackling the above issues, extensive experiments on sim2real benchmarks show that the proposed method outperforms the state-of-the-art methods. Our codes are available at https://github.com/alpc91/TransDA

Jiangmeng Li, Yanan Zhang, Wenwen Qiang et al.

Few-shot learning models learn representations with limited human annotations, and such a learning paradigm demonstrates practicability in various tasks, e.g., image classification, object detection, etc. However, few-shot object detection methods suffer from an intrinsic defect that the limited training data makes the model cannot sufficiently explore semantic information. To tackle this, we introduce knowledge distillation to the few-shot object detection learning paradigm. We further run a motivating experiment, which demonstrates that in the process of knowledge distillation, the empirical error of the teacher model degenerates the prediction performance of the few-shot object detection model as the student. To understand the reasons behind this phenomenon, we revisit the learning paradigm of knowledge distillation on the few-shot object detection task from the causal theoretic standpoint, and accordingly, develop a Structural Causal Model. Following the theoretical guidance, we propose a backdoor adjustment-based knowledge distillation method for the few-shot object detection task, namely Disentangle and Remerge (D&R), to perform conditional causal intervention toward the corresponding Structural Causal Model. Empirically, the experiments on benchmarks demonstrate that D&R can yield significant performance boosts in few-shot object detection. Code is available at https://github.com/ZYN-1101/DandR.git.

Wenbing Huang, Jiaqi Han, Yu Rong et al.

Learning to reason about relations and dynamics over multiple interacting objects is a challenging topic in machine learning. The challenges mainly stem from that the interacting systems are exponentially-compositional, symmetrical, and commonly geometrically-constrained. Current methods, particularly the ones based on equivariant Graph Neural Networks (GNNs), have targeted on the first two challenges but remain immature for constrained systems. In this paper, we propose Graph Mechanics Network (GMN) which is combinatorially efficient, equivariant and constraint-aware. The core of GMN is that it represents, by generalized coordinates, the forward kinematics information (positions and velocities) of a structural object. In this manner, the geometrical constraints are implicitly and naturally encoded in the forward kinematics. Moreover, to allow equivariant message passing in GMN, we have developed a general form of orthogonality-equivariant functions, given that the dynamics of constrained systems are more complicated than the unconstrained counterparts. Theoretically, the proposed equivariant formulation is proved to be universally expressive under certain conditions. Extensive experiments support the advantages of GMN compared to the state-of-the-art GNNs in terms of prediction accuracy, constraint satisfaction and data efficiency on the simulated systems consisting of particles, sticks and hinges, as well as two real-world datasets for molecular dynamics prediction and human motion capture.

Yuhong Deng, Xiaofeng Guo, Yixuan Wei et al.

In this paper, a novel robotic grasping system is established to automatically pick up objects in cluttered scenes. A composite robotic hand composed of a suction cup and a gripper is designed for grasping the object stably. The suction cup is used for lifting the object from the clutter first and the gripper for grasping the object accordingly. We utilize the affordance map to provide pixel-wise lifting point candidates for the suction cup. To obtain a good affordance map, the active exploration mechanism is introduced to the system. An effective metric is designed to calculate the reward for the current affordance map, and a deep Q-Network (DQN) is employed to guide the robotic hand to actively explore the environment until the generated affordance map is suitable for grasping. Experimental results have demonstrated that the proposed robotic grasping system is able to greatly increase the success rate of the robotic grasping in cluttered scenes.

Zhanhao Hu, Siyuan Huang, Xiaopei Zhu et al.

Nowadays, cameras equipped with AI systems can capture and analyze images to detect people automatically. However, the AI system can make mistakes when receiving deliberately designed patterns in the real world, i.e., physical adversarial examples. Prior works have shown that it is possible to print adversarial patches on clothes to evade DNN-based person detectors. However, these adversarial examples could have catastrophic drops in the attack success rate when the viewing angle (i.e., the camera's angle towards the object) changes. To perform a multi-angle attack, we propose Adversarial Texture (AdvTexture). AdvTexture can cover clothes with arbitrary shapes so that people wearing such clothes can hide from person detectors from different viewing angles. We propose a generative method, named Toroidal-Cropping-based Expandable Generative Attack (TC-EGA), to craft AdvTexture with repetitive structures. We printed several pieces of cloth with AdvTexure and then made T-shirts, skirts, and dresses in the physical world. Experiments showed that these clothes could fool person detectors in the physical world.

Yinfeng Yu, Lele Cao, Fuchun Sun et al. · tsinghua

Audio-visual embodied navigation, as a hot research topic, aims training a robot to reach an audio target using egocentric visual (from the sensors mounted on the robot) and audio (emitted from the target) input. The audio-visual information fusion strategy is naturally important to the navigation performance, but the state-of-the-art methods still simply concatenate the visual and audio features, potentially ignoring the direct impact of context. Moreover, the existing approaches requires either phase-wise training or additional aid (e.g. topology graph and sound semantics). Up till this date, the work that deals with the more challenging setup with moving target(s) is still rare. As a result, we propose an end-to-end framework FSAAVN (feature self-attention audio-visual navigation) to learn chasing after a moving audio target using a context-aware audio-visual fusion strategy implemented as a self-attention module. Our thorough experiments validate the superior performance (both quantitatively and qualitatively) of FSAAVN in comparison with the state-of-the-arts, and also provide unique insights about the choice of visual modalities, visual/audio encoder backbones and fusion patterns.

Jiaqi Han, Wenbing Huang, Yu Rong et al.

It has been discovered that Graph Convolutional Networks (GCNs) encounter a remarkable drop in performance when multiple layers are piled up. The main factor that accounts for why deep GCNs fail lies in over-smoothing, which isolates the network output from the input with the increase of network depth, weakening expressivity and trainability. In this paper, we start by investigating refined measures upon DropEdge -- an existing simple yet effective technique to relieve over-smoothing. We term our method as DropEdge++ for its two structure-aware samplers in contrast to DropEdge: layer-dependent sampler and feature-dependent sampler. Regarding the layer-dependent sampler, we interestingly find that increasingly sampling edges from the bottom layer yields superior performance than the decreasing counterpart as well as DropEdge. We theoretically reveal this phenomenon with Mean-Edge-Number (MEN), a metric closely related to over-smoothing. For the feature-dependent sampler, we associate the edge sampling probability with the feature similarity of node pairs, and prove that it further correlates the convergence subspace of the output layer with the input features. Extensive experiments on several node classification benchmarks, including both full- and semi- supervised tasks, illustrate the efficacy of DropEdge++ and its compatibility with a variety of backbones by achieving generally better performance over DropEdge and the no-drop version.

Yinfeng Yu, Changan Chen, Lele Cao et al. · tsinghua

As humans, we hear sound every second of our life. The sound we hear is often affected by the acoustics of the environment surrounding us. For example, a spacious hall leads to more reverberation. Room Impulse Responses (RIR) are commonly used to characterize environment acoustics as a function of the scene geometry, materials, and source/receiver locations. Traditionally, RIRs are measured by setting up a loudspeaker and microphone in the environment for all source/receiver locations, which is time-consuming and inefficient. We propose to let two robots measure the environment's acoustics by actively moving and emitting/receiving sweep signals. We also devise a collaborative multi-agent policy where these two robots are trained to explore the environment's acoustics while being rewarded for wide exploration and accurate prediction. We show that the robots learn to collaborate and move to explore environment acoustics while minimizing the prediction error. To the best of our knowledge, we present the very first problem formulation and solution to the task of collaborative environment acoustics measurements with multiple agents.

Yikai Wang, TengQi Ye, Lele Cao et al.

3D object detection is a crucial research topic in computer vision, which usually uses 3D point clouds as input in conventional setups. Recently, there is a trend of leveraging multiple sources of input data, such as complementing the 3D point cloud with 2D images that often have richer color and fewer noises. However, due to the heterogeneous geometrics of the 2D and 3D representations, it prevents us from applying off-the-shelf neural networks to achieve multimodal fusion. To that end, we propose Bridged Transformer (BrT), an end-to-end architecture for 3D object detection. BrT is simple and effective, which learns to identify 3D and 2D object bounding boxes from both points and image patches. A key element of BrT lies in the utilization of object queries for bridging 3D and 2D spaces, which unifies different sources of data representations in Transformer. We adopt a form of feature aggregation realized by point-to-patch projections which further strengthen the correlations between images and points. Moreover, BrT works seamlessly for fusing the point cloud with multi-view images. We experimentally show that BrT surpasses state-of-the-art methods on SUN RGB-D and ScanNetV2 datasets.

Yikai Wang, Yinpeng Dong, Fuchun Sun et al.

This work focuses on the 3D reconstruction of non-rigid objects based on monocular RGB video sequences. Concretely, we aim at building high-fidelity models for generic object categories and casually captured scenes. To this end, we do not assume known root poses of objects, and do not utilize category-specific templates or dense pose priors. The key idea of our method, Root Pose Decomposition (RPD), is to maintain a per-frame root pose transformation, meanwhile building a dense field with local transformations to rectify the root pose. The optimization of local transformations is performed by point registration to the canonical space. We also adapt RPD to multi-object scenarios with object occlusions and individual differences. As a result, RPD allows non-rigid 3D reconstruction for complicated scenarios containing objects with large deformations, complex motion patterns, occlusions, and scale diversities of different individuals. Such a pipeline potentially scales to diverse sets of objects in the wild. We experimentally show that RPD surpasses state-of-the-art methods on the challenging DAVIS, OVIS, and AMA datasets.

Tianying Ji, Yu Luo, Fuchun Sun et al.

Learning high-quality $Q$-value functions plays a key role in the success of many modern off-policy deep reinforcement learning (RL) algorithms. Previous works primarily focus on addressing the value overestimation issue, an outcome of adopting function approximators and off-policy learning. Deviating from the common viewpoint, we observe that $Q$-values are often underestimated in the latter stage of the RL training process, potentially hindering policy learning and reducing sample efficiency. We find that such a long-neglected phenomenon is often related to the use of inferior actions from the current policy in Bellman updates as compared to the more optimal action samples in the replay buffer. To address this issue, our insight is to incorporate sufficient exploitation of past successes while maintaining exploration optimism. We propose the Blended Exploitation and Exploration (BEE) operator, a simple yet effective approach that updates $Q$-value using both historical best-performing actions and the current policy. Based on BEE, the resulting practical algorithm BAC outperforms state-of-the-art methods in over 50 continuous control tasks and achieves strong performance in failure-prone scenarios and real-world robot tasks. Benchmark results and videos are available at https://jity16.github.io/BEE/.

Yikai Wang, Wenbing Huang, Yinpeng Dong et al.

Binary Neural Network (BNN) represents convolution weights with 1-bit values, which enhances the efficiency of storage and computation. This paper is motivated by a previously revealed phenomenon that the binary kernels in successful BNNs are nearly power-law distributed: their values are mostly clustered into a small number of codewords. This phenomenon encourages us to compact typical BNNs and obtain further close performance through learning non-repetitive kernels within a binary kernel subspace. Specifically, we regard the binarization process as kernel grouping in terms of a binary codebook, and our task lies in learning to select a smaller subset of codewords from the full codebook. We then leverage the Gumbel-Sinkhorn technique to approximate the codeword selection process, and develop the Permutation Straight-Through Estimator (PSTE) that is able to not only optimize the selection process end-to-end but also maintain the non-repetitive occupancy of selected codewords. Experiments verify that our method reduces both the model size and bit-wise computational costs, and achieves accuracy improvements compared with state-of-the-art BNNs under comparable budgets.

Tianying Ji, Yu Luo, Fuchun Sun et al.

Designing and analyzing model-based RL (MBRL) algorithms with guaranteed monotonic improvement has been challenging, mainly due to the interdependence between policy optimization and model learning. Existing discrepancy bounds generally ignore the impacts of model shifts, and their corresponding algorithms are prone to degrade performance by drastic model updating. In this work, we first propose a novel and general theoretical scheme for a non-decreasing performance guarantee of MBRL. Our follow-up derived bounds reveal the relationship between model shifts and performance improvement. These discoveries encourage us to formulate a constrained lower-bound optimization problem to permit the monotonicity of MBRL. A further example demonstrates that learning models from a dynamically-varying number of explorations benefit the eventual returns. Motivated by these analyses, we design a simple but effective algorithm CMLO (Constrained Model-shift Lower-bound Optimization), by introducing an event-triggered mechanism that flexibly determines when to update the model. Experiments show that CMLO surpasses other state-of-the-art methods and produces a boost when various policy optimization methods are employed.

Xiao Li, Ziqi Wang, Bo Zhang et al.

Adversarial attacks can easily fool object recognition systems based on deep neural networks (DNNs). Although many defense methods have been proposed in recent years, most of them can still be adaptively evaded. One reason for the weak adversarial robustness may be that DNNs are only supervised by category labels and do not have part-based inductive bias like the recognition process of humans. Inspired by a well-known theory in cognitive psychology -- recognition-by-components, we propose a novel object recognition model ROCK (Recognizing Object by Components with human prior Knowledge). It first segments parts of objects from images, then scores part segmentation results with predefined human prior knowledge, and finally outputs prediction based on the scores. The first stage of ROCK corresponds to the process of decomposing objects into parts in human vision. The second stage corresponds to the decision process of the human brain. ROCK shows better robustness than classical recognition models across various attack settings. These results encourage researchers to rethink the rationality of currently widely-used DNN-based object recognition models and explore the potential of part-based models, once important but recently ignored, for improving robustness.

Xueyi Liu, Yu Rong, Tingyang Xu et al.

Graph instance contrastive learning has been proved as an effective task for Graph Neural Network (GNN) pre-training. However, one key issue may seriously impede the representative power in existing works: Positive instances created by current methods often miss crucial information of graphs or even yield illegal instances (such as non-chemically-aware graphs in molecular generation). To remedy this issue, we propose to select positive graph instances directly from existing graphs in the training set, which ultimately maintains the legality and similarity to the target graphs. Our selection is based on certain domain-specific pair-wise similarity measurements as well as sampling from a hierarchical graph encoding similarity relations among graphs. Besides, we develop an adaptive node-level pre-training method to dynamically mask nodes to distribute them evenly in the graph. We conduct extensive experiments on $13$ graph classification and node classification benchmark datasets from various domains. The results demonstrate that the GNN models pre-trained by our strategies can outperform those trained-from-scratch models as well as the variants obtained by existing methods.

Hang Gao, Jiangmeng Li, Wenwen Qiang et al.

The prevailing graph neural network models have achieved significant progress in graph representation learning. However, in this paper, we uncover an ever-overlooked phenomenon: the pre-trained graph representation learning model tested with full graphs underperforms the model tested with well-pruned graphs. This observation reveals that there exist confounders in graphs, which may interfere with the model learning semantic information, and current graph representation learning methods have not eliminated their influence. To tackle this issue, we propose Robust Causal Graph Representation Learning (RCGRL) to learn robust graph representations against confounding effects. RCGRL introduces an active approach to generate instrumental variables under unconditional moment restrictions, which empowers the graph representation learning model to eliminate confounders, thereby capturing discriminative information that is causally related to downstream predictions. We offer theorems and proofs to guarantee the theoretical effectiveness of the proposed approach. Empirically, we conduct extensive experiments on a synthetic dataset and multiple benchmark datasets. The results demonstrate that compared with state-of-the-art methods, RCGRL achieves better prediction performance and generalization ability.

Hang Gao, Jiangmeng Li, Wenwen Qiang et al.

Benefiting from the injection of human prior knowledge, graphs, as derived discrete data, are semantically dense so that models can efficiently learn the semantic information from such data. Accordingly, graph neural networks (GNNs) indeed achieve impressive success in various fields. Revisiting the GNN learning paradigms, we discover that the relationship between human expertise and the knowledge modeled by GNNs still confuses researchers. To this end, we introduce motivating experiments and derive an empirical observation that the GNNs gradually learn human expertise in general domains. By further observing the ramifications of introducing expertise logic into graph representation learning, we conclude that leading the GNNs to learn human expertise can improve the model performance. Hence, we propose a novel graph representation learning method to incorporate human expert knowledge into GNN models. The proposed method ensures that the GNN model can not only acquire the expertise held by human experts but also engage in end-to-end learning from datasets. Plentiful experiments on the crafted and real-world domains support the consistent effectiveness of the proposed method.

Dazhao Du, Enhan Li, Lingyu Si et al.

Video action segmentation under timestamp supervision has recently received much attention due to lower annotation costs. Most existing methods generate pseudo-labels for all frames in each video to train the segmentation model. However, these methods suffer from incorrect pseudo-labels, especially for the semantically unclear frames in the transition region between two consecutive actions, which we call ambiguous intervals. To address this issue, we propose a novel framework from the perspective of clustering, which includes the following two parts. First, pseudo-label ensembling generates incomplete but high-quality pseudo-label sequences, where the frames in ambiguous intervals have no pseudo-labels. Second, iterative clustering iteratively propagates the pseudo-labels to the ambiguous intervals by clustering, and thus updates the pseudo-label sequences to train the model. We further introduce a clustering loss, which encourages the features of frames within the same action segment more compact. Extensive experiments show the effectiveness of our method.

Lingyu Si, Hongwei Dong, Wenwen Qiang et al.

Due to limitations in data quality, some essential visual tasks are difficult to perform independently. Introducing previously unavailable information to transfer informative dark knowledge has been a common way to solve such hard tasks. However, research on why transferred knowledge works has not been extensively explored. To address this issue, in this paper, we discover the correlation between feature discriminability and dimensional structure (DS) by analyzing and observing features extracted from simple and hard tasks. On this basis, we express DS using deep channel-wise correlation and intermediate spatial distribution, and propose a novel cross-modal knowledge distillation (CMKD) method for better supervised cross-modal learning (CML) performance. The proposed method enforces output features to be channel-wise independent and intermediate ones to be uniformly distributed, thereby learning semantically irrelevant features from the hard task to boost its accuracy. This is especially useful in specific applications where the performance gap between dual modalities is relatively large. Furthermore, we collect a real-world CML dataset to promote community development. The dataset contains more than 10,000 paired optical and radar images and is continuously being updated. Experimental results on real-world and benchmark datasets validate the effectiveness of the proposed method.

Zeen Song, Xingzhe Su, Jingyao Wang et al.

In recent years, self-supervised learning (SSL) has emerged as a promising approach for extracting valuable representations from unlabeled data. One successful SSL method is contrastive learning, which aims to bring positive examples closer while pushing negative examples apart. Many current contrastive learning approaches utilize a parameterized projection head. Through a combination of empirical analysis and theoretical investigation, we provide insights into the internal mechanisms of the projection head and its relationship with the phenomenon of dimensional collapse. Our findings demonstrate that the projection head enhances the quality of representations by performing contrastive loss in a projected subspace. Therefore, we propose an assumption that only a subset of features is necessary when minimizing the contrastive loss of a mini-batch of data. Theoretical analysis further suggests that a sparse projection head can enhance generalization, leading us to introduce SparseHead - a regularization term that effectively constrains the sparsity of the projection head, and can be seamlessly integrated with any self-supervised learning (SSL) approaches. Our experimental results validate the effectiveness of SparseHead, demonstrating its ability to improve the performance of existing contrastive methods.

Zeen Song, Wenwen Qiang, Changwen Zheng et al.

Self-supervised learning (SSL) has recently shown notable success in various visual tasks. However, in terms of discriminability, SSL is still not on par with supervised learning (SL). This paper identifies a key issue, the ``crowding problem," where features from different classes are not well-separated, and there is high intra-class variance. In contrast, SL ensures clear class separation. Our analysis reveals that SSL objectives do not adequately constrain the relationships between samples and their augmentations, leading to poorer performance in complex tasks. We further establish a theoretical framework that connects SSL objectives to cross-entropy risk bounds, explaining how reducing intra-class variance and increasing inter-class separation can improve generalization. To address this, we propose the Dynamic Semantic Adjuster (DSA), a learnable regulator that enhances feature aggregation and separation while being robust to outliers. Comprehensive experiments conducted on diverse benchmark datasets validate that DSA leads to substantial gains in SSL performance, narrowing the performance gap with SL.

Pengkun Liu, Yikai Wang, Fuchun Sun et al.

Encouraged by the growing availability of pre-trained 2D diffusion models, image-to-3D generation by leveraging Score Distillation Sampling (SDS) is making remarkable progress. Most existing methods combine novel-view lifting from 2D diffusion models which usually take the reference image as a condition while applying hard L2 image supervision at the reference view. Yet heavily adhering to the image is prone to corrupting the inductive knowledge of the 2D diffusion model leading to flat or distorted 3D generation frequently. In this work, we reexamine image-to-3D in a novel perspective and present Isotropic3D, an image-to-3D generation pipeline that takes only an image CLIP embedding as input. Isotropic3D allows the optimization to be isotropic w.r.t. the azimuth angle by solely resting on the SDS loss. The core of our framework lies in a two-stage diffusion model fine-tuning. Firstly, we fine-tune a text-to-3D diffusion model by substituting its text encoder with an image encoder, by which the model preliminarily acquires image-to-image capabilities. Secondly, we perform fine-tuning using our Explicit Multi-view Attention (EMA) which combines noisy multi-view images with the noise-free reference image as an explicit condition. CLIP embedding is sent to the diffusion model throughout the whole process while reference images are discarded once after fine-tuning. As a result, with a single image CLIP embedding, Isotropic3D is capable of generating multi-view mutually consistent images and also a 3D model with more symmetrical and neat content, well-proportioned geometry, rich colored texture, and less distortion compared with existing image-to-3D methods while still preserving the similarity to the reference image to a large extent. The project page is available at https://isotropic3d.github.io/. The code and models are available at https://github.com/pkunliu/Isotropic3D.

Jiangmeng Li, Yifan Jin, Hang Gao et al.

Graph contrastive learning (GCL) aims to align the positive features while differentiating the negative features in the latent space by minimizing a pair-wise contrastive loss. As the embodiment of an outstanding discriminative unsupervised graph representation learning approach, GCL achieves impressive successes in various graph benchmarks. However, such an approach falls short of recognizing the topology isomorphism of graphs, resulting in that graphs with relatively homogeneous node features cannot be sufficiently discriminated. By revisiting classic graph topology recognition works, we disclose that the corresponding expertise intuitively complements GCL methods. To this end, we propose a novel hierarchical topology isomorphism expertise embedded graph contrastive learning, which introduces knowledge distillations to empower GCL models to learn the hierarchical topology isomorphism expertise, including the graph-tier and subgraph-tier. On top of this, the proposed method holds the feature of plug-and-play, and we empirically demonstrate that the proposed method is universal to multiple state-of-the-art GCL models. The solid theoretical analyses are further provided to prove that compared with conventional GCL methods, our method acquires the tighter upper bound of Bayes classification error. We conduct extensive experiments on real-world benchmarks to exhibit the performance superiority of our method over candidate GCL methods, e.g., for the real-world graph representation learning experiments, the proposed method beats the state-of-the-art method by 0.23% on unsupervised representation learning setting, 0.43% on transfer learning setting. Our code is available at https://github.com/jyf123/HTML.

Runfa Chen, Ling Wang, Yu Du et al.

Learning policies for multi-entity systems in 3D environments is far more complicated against single-entity scenarios, due to the exponential expansion of the global state space as the number of entities increases. One potential solution of alleviating the exponential complexity is dividing the global space into independent local views that are invariant to transformations including translations and rotations. To this end, this paper proposes Subequivariant Hierarchical Neural Networks (SHNN) to facilitate multi-entity policy learning. In particular, SHNN first dynamically decouples the global space into local entity-level graphs via task assignment. Second, it leverages subequivariant message passing over the local entity-level graphs to devise local reference frames, remarkably compressing the representation redundancy, particularly in gravity-affected environments. Furthermore, to overcome the limitations of existing benchmarks in capturing the subtleties of multi-entity systems under the Euclidean symmetry, we propose the Multi-entity Benchmark (MEBEN), a new suite of environments tailored for exploring a wide range of multi-entity reinforcement learning. Extensive experiments demonstrate significant advancements of SHNN on the proposed benchmarks compared to existing methods. Comprehensive ablations are conducted to verify the indispensability of task assignment and subequivariance.

He Liu, Yikai Wang, Huaping Liu et al.

Data-free knowledge distillation is able to utilize the knowledge learned by a large teacher network to augment the training of a smaller student network without accessing the original training data, avoiding privacy, security, and proprietary risks in real applications. In this line of research, existing methods typically follow an inversion-and-distillation paradigm in which a generative adversarial network on-the-fly trained with the guidance of the pre-trained teacher network is used to synthesize a large-scale sample set for knowledge distillation. In this paper, we reexamine this common data-free knowledge distillation paradigm, showing that there is considerable room to improve the overall training efficiency through a lens of ``small-scale inverted data for knowledge distillation". In light of three empirical observations indicating the importance of how to balance class distributions in terms of synthetic sample diversity and difficulty during both data inversion and distillation processes, we propose Small Scale Data-free Knowledge Distillation SSD-KD. In formulation, SSD-KD introduces a modulating function to balance synthetic samples and a priority sampling function to select proper samples, facilitated by a dynamic replay buffer and a reinforcement learning strategy. As a result, SSD-KD can perform distillation training conditioned on an extremely small scale of synthetic samples (e.g., 10X less than the original training data scale), making the overall training efficiency one or two orders of magnitude faster than many mainstream methods while retaining superior or competitive model performance, as demonstrated on popular image classification and semantic segmentation benchmarks. The code is available at https://github.com/OSVAI/SSD-KD.

Ruichu Cai, Junxian Huang, Zhenhui Yang et al.

Time series domain adaptation aims to transfer the complex temporal dependence from the labeled source domain to the unlabeled target domain. Recent advances leverage the stable causal mechanism over observed variables to model the domain-invariant temporal dependence. However, modeling precise causal structures in high-dimensional data, such as videos, remains challenging. Additionally, direct causal edges may not exist among observed variables (e.g., pixels). These limitations hinder the applicability of existing approaches to real-world scenarios. To address these challenges, we find that the high-dimension time series data are generated from the low-dimension latent variables, which motivates us to model the causal mechanisms of the temporal latent process. Based on this intuition, we propose a latent causal mechanism identification framework that guarantees the uniqueness of the reconstructed latent causal structures. Specifically, we first identify latent variables by utilizing sufficient changes in historical information. Moreover, by enforcing the sparsity of the relationships of latent variables, we can achieve identifiable latent causal structures. Built on the theoretical results, we develop the Latent Causality Alignment (LCA) model that leverages variational inference, which incorporates an intra-domain latent sparsity constraint for latent structure reconstruction and an inter-domain latent sparsity constraint for domain-invariant structure reconstruction. Experiment results on eight benchmarks show a general improvement in the domain-adaptive time series classification and forecasting tasks, highlighting the effectiveness of our method in real-world scenarios. Codes are available at https://github.com/DMIRLAB-Group/LCA.

Ruichu Cai, Siyang Huang, Jie Qiao et al.

As a key component to intuitive cognition and reasoning solutions in human intelligence, causal knowledge provides great potential for reinforcement learning (RL) agents' interpretability towards decision-making by helping reduce the searching space. However, there is still a considerable gap in discovering and incorporating causality into RL, which hinders the rapid development of causal RL. In this paper, we consider explicitly modeling the generation process of states with the causal graphical model, based on which we augment the policy. We formulate the causal structure updating into the RL interaction process with active intervention learning of the environment. To optimize the derived objective, we propose a framework with theoretical performance guarantees that alternates between two steps: using interventions for causal structure learning during exploration and using the learned causal structure for policy guidance during exploitation. Due to the lack of public benchmarks that allow direct intervention in the state space, we design the root cause localization task in our simulated fault alarm environment and then empirically show the effectiveness and robustness of the proposed method against state-of-the-art baselines. Theoretical analysis shows that our performance improvement attributes to the virtuous cycle of causal-guided policy learning and causal structure learning, which aligns with our experimental results. Codes are available at https://github.com/DMIRLAB-Group/FaultAlarm_RL.

Yikai Wang, Fuchun Sun, Wenbing Huang et al.

Multimodal fusion and multitask learning are two vital topics in machine learning. Despite the fruitful progress, existing methods for both problems are still brittle to the same challenge -- it remains dilemmatic to integrate the common information across modalities (resp. tasks) meanwhile preserving the specific patterns of each modality (resp. task). Besides, while they are actually closely related to each other, multimodal fusion and multitask learning are rarely explored within the same methodological framework before. In this paper, we propose Channel-Exchanging-Network (CEN) which is self-adaptive, parameter-free, and more importantly, applicable for multimodal and multitask dense image prediction. At its core, CEN adaptively exchanges channels between subnetworks of different modalities. Specifically, the channel exchanging process is self-guided by individual channel importance that is measured by the magnitude of Batch-Normalization (BN) scaling factor during training. For the application of dense image prediction, the validity of CEN is tested by four different scenarios: multimodal fusion, cycle multimodal fusion, multitask learning, and multimodal multitask learning. Extensive experiments on semantic segmentation via RGB-D data and image translation through multi-domain input verify the effectiveness of CEN compared to state-of-the-art methods. Detailed ablation studies have also been carried out, which demonstrate the advantage of each component we propose. Our code is available at https://github.com/yikaiw/CEN.

Yikai Wang, Yi Yang, Fuchun Sun et al.

In the low-bit quantization field, training Binary Neural Networks (BNNs) is the extreme solution to ease the deployment of deep models on resource-constrained devices, having the lowest storage cost and significantly cheaper bit-wise operations compared to 32-bit floating-point counterparts. In this paper, we introduce Sub-bit Neural Networks (SNNs), a new type of binary quantization design tailored to compress and accelerate BNNs. SNNs are inspired by an empirical observation, showing that binary kernels learnt at convolutional layers of a BNN model are likely to be distributed over kernel subsets. As a result, unlike existing methods that binarize weights one by one, SNNs are trained with a kernel-aware optimization framework, which exploits binary quantization in the fine-grained convolutional kernel space. Specifically, our method includes a random sampling step generating layer-specific subsets of the kernel space, and a refinement step learning to adjust these subsets of binary kernels via optimization. Experiments on visual recognition benchmarks and the hardware deployment on FPGA validate the great potentials of SNNs. For instance, on ImageNet, SNNs of ResNet-18/ResNet-34 with 0.56-bit weights achieve 3.13/3.33 times runtime speed-up and 1.8 times compression over conventional BNNs with moderate drops in recognition accuracy. Promising results are also obtained when applying SNNs to binarize both weights and activations. Our code is available at https://github.com/yikaiw/SNN.

Yikai Wang, Wenbing Huang, Fuchun Sun et al.

Deep multimodal fusion by using multiple sources of data for classification or regression has exhibited a clear advantage over the unimodal counterpart on various applications. Yet, current methods including aggregation-based and alignment-based fusion are still inadequate in balancing the trade-off between inter-modal fusion and intra-modal processing, incurring a bottleneck of performance improvement. To this end, this paper proposes Channel-Exchanging-Network (CEN), a parameter-free multimodal fusion framework that dynamically exchanges channels between sub-networks of different modalities. Specifically, the channel exchanging process is self-guided by individual channel importance that is measured by the magnitude of Batch-Normalization (BN) scaling factor during training. The validity of such exchanging process is also guaranteed by sharing convolutional filters yet keeping separate BN layers across modalities, which, as an add-on benefit, allows our multimodal architecture to be almost as compact as a unimodal network. Extensive experiments on semantic segmentation via RGB-D data and image translation through multi-domain input verify the effectiveness of our CEN compared to current state-of-the-art methods. Detailed ablation studies have also been carried out, which provably affirm the advantage of each component we propose. Our code is available at https://github.com/yikaiw/CEN.

Yikai Wang, Fuchun Sun, Duo Li et al.

We propose a general method to train a single convolutional neural network which is capable of switching image resolutions at inference. Thus the running speed can be selected to meet various computational resource limits. Networks trained with the proposed method are named Resolution Switchable Networks (RS-Nets). The basic training framework shares network parameters for handling images which differ in resolution, yet keeps separate batch normalization layers. Though it is parameter-efficient in design, it leads to inconsistent accuracy variations at different resolutions, for which we provide a detailed analysis from the aspect of the train-test recognition discrepancy. A multi-resolution ensemble distillation is further designed, where a teacher is learnt on the fly as a weighted ensemble over resolutions. Thanks to the ensemble and knowledge distillation, RS-Nets enjoy accuracy improvements at a wide range of resolutions compared with individually trained models. Extensive experiments on the ImageNet dataset are provided, and we additionally consider quantization problems. Code and models are available at https://github.com/yikaiw/RS-Nets.

Runfa Chen, Wenbing Huang, Binghui Huang et al.

Unsupervised image-to-image translation is a central task in computer vision. Current translation frameworks will abandon the discriminator once the training process is completed. This paper contends a novel role of the discriminator by reusing it for encoding the images of the target domain. The proposed architecture, termed as NICE-GAN, exhibits two advantageous patterns over previous approaches: First, it is more compact since no independent encoding component is required; Second, this plug-in encoder is directly trained by the adversary loss, making it more informative and trained more effectively if a multi-scale discriminator is applied. The main issue in NICE-GAN is the coupling of translation with discrimination along the encoder, which could incur training inconsistency when we play the min-max game via GAN. To tackle this issue, we develop a decoupled training strategy by which the encoder is only trained when maximizing the adversary loss while keeping frozen otherwise. Extensive experiments on four popular benchmarks demonstrate the superior performance of NICE-GAN over state-of-the-art methods in terms of FID, KID, and also human preference. Comprehensive ablation studies are also carried out to isolate the validity of each proposed component. Our codes are available at https://github.com/alpc91/NICE-GAN-pytorch.

Luxuan Li, Tao Kong, Fuchun Sun et al.

Detecting actions in videos is an important yet challenging task. Previous works usually utilize (a) sliding window paradigms, or (b) per-frame action scoring and grouping to enumerate the possible temporal locations. Their performances are also limited to the designs of sliding windows or grouping strategies. In this paper, we present a simple and effective method for temporal action proposal generation, named Deep Point-wise Prediction (DPP). DPP simultaneously predicts the action existing possibility and the corresponding temporal locations, without the utilization of any handcrafted sliding window or grouping. The whole system is end-to-end trained with joint loss of temporal action proposal classification and location prediction. We conduct extensive experiments to verify its effectiveness, generality and robustness on standard THUMOS14 dataset. DPP runs more than 1000 frames per second, which largely satisfies the real-time requirement. The code is available at https://github.com/liluxuan1997/DPP.

Xinzhou Wang, Yikai Wang, Junliang Ye et al.

Advances in 3D generation have facilitated sequential 3D model generation (a.k.a 4D generation), yet its application for animatable objects with large motion remains scarce. Our work proposes AnimatableDreamer, a text-to-4D generation framework capable of generating diverse categories of non-rigid objects on skeletons extracted from a monocular video. At its core, AnimatableDreamer is equipped with our novel optimization design dubbed Canonical Score Distillation (CSD), which lifts 2D diffusion for temporal consistent 4D generation. CSD, designed from a score gradient perspective, generates a canonical model with warp-robustness across different articulations. Notably, it also enhances the authenticity of bones and skinning by integrating inductive priors from a diffusion model. Furthermore, with multi-view distillation, CSD infers invisible regions, thereby improving the fidelity of monocular non-rigid reconstruction. Extensive experiments demonstrate the capability of our method in generating high-flexibility text-guided 3D models from the monocular video, while also showing improved reconstruction performance over existing non-rigid reconstruction methods.

Kai Sun, Siyan Xue, Fuchun Sun et al.

Recent advancements in deep learning have significantly revolutionized the field of clinical diagnosis and treatment, offering novel approaches to improve diagnostic precision and treatment efficacy across diverse clinical domains, thus driving the pursuit of precision medicine. The growing availability of multi-organ and multimodal datasets has accelerated the development of large-scale Medical Multimodal Foundation Models (MMFMs). These models, known for their strong generalization capabilities and rich representational power, are increasingly being adapted to address a wide range of clinical tasks, from early diagnosis to personalized treatment strategies. This review offers a comprehensive analysis of recent developments in MMFMs, focusing on three key aspects: datasets, model architectures, and clinical applications. We also explore the challenges and opportunities in optimizing multimodal representations and discuss how these advancements are shaping the future of healthcare by enabling improved patient outcomes and more efficient clinical workflows.

Tianying Ji, Yongyuan Liang, Yan Zeng et al. · tsinghua

The varying significance of distinct primitive behaviors during the policy learning process has been overlooked by prior model-free RL algorithms. Leveraging this insight, we explore the causal relationship between different action dimensions and rewards to evaluate the significance of various primitive behaviors during training. We introduce a causality-aware entropy term that effectively identifies and prioritizes actions with high potential impacts for efficient exploration. Furthermore, to prevent excessive focus on specific primitive behaviors, we analyze the gradient dormancy phenomenon and introduce a dormancy-guided reset mechanism to further enhance the efficacy of our method. Our proposed algorithm, ACE: Off-policy Actor-critic with Causality-aware Entropy regularization, demonstrates a substantial performance advantage across 29 diverse continuous control tasks spanning 7 domains compared to model-free RL baselines, which underscores the effectiveness, versatility, and efficient sample efficiency of our approach. Benchmark results and videos are available at https://ace-rl.github.io/.

Dazhao Du, Enhan Li, Lingyu Si et al.

Underwater image enhancement (UIE) aims to generate clear images from low-quality underwater images. Due to the unavailability of clear reference images, researchers often synthesize them to construct paired datasets for training deep models. However, these synthesized images may sometimes lack quality, adversely affecting training outcomes. To address this issue, we propose UIE with Diffusion Prior (UIEDP), a novel framework treating UIE as a posterior distribution sampling process of clear images conditioned on degraded underwater inputs. Specifically, UIEDP combines a pre-trained diffusion model capturing natural image priors with any existing UIE algorithm, leveraging the latter to guide conditional generation. The diffusion prior mitigates the drawbacks of inferior synthetic images, resulting in higher-quality image generation. Extensive experiments have demonstrated that our UIEDP yields significant improvements across various metrics, especially no-reference image quality assessment. And the generated enhanced images also exhibit a more natural appearance.

Lei Lv, Yunfei Li, Yu Luo et al.

We present \textbf{FlowRL}, a novel framework for online reinforcement learning that integrates flow-based policy representation with Wasserstein-2-regularized optimization. We argue that in addition to training signals, enhancing the expressiveness of the policy class is crucial for the performance gains in RL. Flow-based generative models offer such potential, excelling at capturing complex, multimodal action distributions. However, their direct application in online RL is challenging due to a fundamental objective mismatch: standard flow training optimizes for static data imitation, while RL requires value-based policy optimization through a dynamic buffer, leading to difficult optimization landscapes. FlowRL first models policies via a state-dependent velocity field, generating actions through deterministic ODE integration from noise. We derive a constrained policy search objective that jointly maximizes Q through the flow policy while bounding the Wasserstein-2 distance to a behavior-optimal policy implicitly derived from the replay buffer. This formulation effectively aligns the flow optimization with the RL objective, enabling efficient and value-aware policy learning despite the complexity of the policy class. Empirical evaluations on DMControl and Humanoidbench demonstrate that FlowRL achieves competitive performance in online reinforcement learning benchmarks.

Zhaoxun Ju, Chao Yang, Hongbo Wang et al.

Language-conditioned robot behavior plays a vital role in executing complex tasks by associating human commands or instructions with perception and actions. The ability to compose long-horizon tasks based on unconstrained language instructions necessitates the acquisition of a diverse set of general-purpose skills. However, acquiring inherent primitive skills in a coupled and long-horizon environment without external rewards or human supervision presents significant challenges. In this paper, we evaluate the relationship between skills and language instructions from a mathematical perspective, employing two forms of mutual information within the framework of language-conditioned policy learning. To maximize the mutual information between language and skills in an unsupervised manner, we propose an end-to-end imitation learning approach known as Language Conditioned Skill Discovery (LCSD). Specifically, we utilize vector quantization to learn discrete latent skills and leverage skill sequences of trajectories to reconstruct high-level semantic instructions. Through extensive experiments on language-conditioned robotic navigation and manipulation tasks, encompassing BabyAI, LORel, and CALVIN, we demonstrate the superiority of our method over prior works. Our approach exhibits enhanced generalization capabilities towards unseen tasks, improved skill interpretability, and notably higher rates of task completion success.

Ling Wang, Runfa Chen, Yikai Wang et al.

Unsupervised non-rigid point cloud shape correspondence underpins a multitude of 3D vision tasks, yet itself is non-trivial given the exponential complexity stemming from inter-point degree-of-freedom, i.e., pose transformations. Based on the assumption of local rigidity, one solution for reducing complexity is to decompose the overall shape into independent local regions using Local Reference Frames (LRFs) that are invariant to SE(3) transformations. However, the focus solely on local structure neglects global geometric contexts, resulting in less distinctive LRFs that lack crucial semantic information necessary for effective matching. Furthermore, such complexity introduces out-of-distribution geometric contexts during inference, thus complicating generalization. To this end, we introduce 1) EquiShape, a novel structure tailored to learn pair-wise LRFs with global structural cues for both spatial and semantic consistency, and 2) LRF-Refine, an optimization strategy generally applicable to LRF-based methods, aimed at addressing the generalization challenges. Specifically, for EquiShape, we employ cross-talk within separate equivariant graph neural networks (Cross-GVP) to build long-range dependencies to compensate for the lack of semantic information in local structure modeling, deducing pair-wise independent SE(3)-equivariant LRF vectors for each point. For LRF-Refine, the optimization adjusts LRFs within specific contexts and knowledge, enhancing the geometric and semantic generalizability of point features. Our overall framework surpasses the state-of-the-art methods by a large margin on three benchmarks. Code and models will be publicly available.

Yi Wang, Yinfeng Yu, Fuchun Sun et al.

Audio-Visual Embodied Navigation aims to enable agents to autonomously navigate to sound sources in unknown 3D environments using auditory cues. While current AVN methods excel on in-distribution sound sources, they exhibit poor cross-source generalization: navigation success rates plummet and search paths become excessively long when agents encounter unheard sounds or unseen environments. This limitation stems from the lack of explicit alignment mechanisms between auditory signals and corresponding visual regions. Policies tend to memorize spurious \enquote{acoustic fingerprint-scenario} correlations during training, leading to blind exploration when exposed to novel sound sources. To address this, we propose the AGVP framework, which transforms sound from policy-memorable acoustic fingerprint cues into spatial guidance. The framework first extracts global auditory context via audio self-attention, then uses this context as queries to guide visual feature attention, highlighting sound-source-related regions at the feature level. Subsequent temporal modeling and policy optimization are then performed. This design, centered on interpretable cross-modal alignment and region reweighting, reduces dependency on specific acoustic fingerprints. Experimental results demonstrate that AGVP improves both navigation efficiency and robustness while achieving superior cross-scenario generalization on previously unheard sounds.

Tao Zhu, Yinfeng Yu, Liejun Wang et al.

Diffusion models have demonstrated remarkable performance in speech synthesis, but typically require multi-step sampling, resulting in low inference efficiency. Recent studies address this issue by distilling diffusion models into consistency models, enabling efficient one-step generation. However, these approaches introduce additional training costs and rely heavily on the performance of pre-trained teacher models. In this paper, we propose ECTSpeech, a simple and effective one-step speech synthesis framework that, for the first time, incorporates the Easy Consistency Tuning (ECT) strategy into speech synthesis. By progressively tightening consistency constraints on a pre-trained diffusion model, ECTSpeech achieves high-quality one-step generation while significantly reducing training complexity. In addition, we design a multi-scale gate module (MSGate) to enhance the denoiser's ability to fuse features at different scales. Experimental results on the LJSpeech dataset demonstrate that ECTSpeech achieves audio quality comparable to state-of-the-art methods under single-step sampling, while substantially reducing the model's training cost and complexity.

Tianheng Zhu, Yinfeng Yu, Liejun Wang et al.

This paper presents EGSTalker, a real-time audio-driven talking head generation framework based on 3D Gaussian Splatting (3DGS). Designed to enhance both speed and visual fidelity, EGSTalker requires only 3-5 minutes of training video to synthesize high-quality facial animations. The framework comprises two key stages: static Gaussian initialization and audio-driven deformation. In the first stage, a multi-resolution hash triplane and a Kolmogorov-Arnold Network (KAN) are used to extract spatial features and construct a compact 3D Gaussian representation. In the second stage, we propose an Efficient Spatial-Audio Attention (ESAA) module to fuse audio and spatial cues, while KAN predicts the corresponding Gaussian deformations. Extensive experiments demonstrate that EGSTalker achieves rendering quality and lip-sync accuracy comparable to state-of-the-art methods, while significantly outperforming them in inference speed. These results highlight EGSTalker's potential for real-time multimedia applications.

Jia Li, Yinfeng Yu, Liejun Wang et al.

In audio-visual navigation (AVN) tasks, an embodied agent must autonomously localize a sound source in unknown and complex 3D environments based on audio-visual signals. Existing methods often rely on static modality fusion strategies and neglect the spatial cues embedded in stereo audio, leading to performance degradation in cluttered or occluded scenes. To address these issues, we propose an end-to-end reinforcement learning-based AVN framework with two key innovations: (1) a \textbf{S}tereo-Aware \textbf{A}ttention \textbf{M}odule (\textbf{SAM}), which learns and exploits the spatial disparity between left and right audio channels to enhance directional sound perception; and (2) an \textbf{A}udio-\textbf{G}uided \textbf{D}ynamic \textbf{F}usion Module (\textbf{AGDF}), which dynamically adjusts the fusion ratio between visual and auditory features based on audio cues, thereby improving robustness to environmental changes. Extensive experiments are conducted on two realistic 3D scene datasets, Replica and Matterport3D, demonstrating that our method significantly outperforms existing approaches in terms of navigation success rate and path efficiency. Notably, our model achieves over 40\% improvement under audio-only conditions compared to the best-performing baselines. These results highlight the importance of explicitly modeling spatial cues from stereo channels and performing deep multi-modal fusion for robust and efficient audio-visual navigation.

Hailong Zhang, Yinfeng Yu, Liejun Wang et al.

Intelligent agents often require collaborative strategies to achieve complex tasks beyond individual capabilities in real-world scenarios. While existing audio-visual navigation (AVN) research mainly focuses on single-agent systems, their limitations emerge in dynamic 3D environments where rapid multi-agent coordination is critical, especially for time-sensitive applications like emergency response. This paper introduces MASTAVN (Multi-Agent Scalable Transformer Audio-Visual Navigation), a scalable framework enabling two agents to collaboratively localize and navigate toward an audio target in shared 3D environments. By integrating cross-agent communication protocols and joint audio-visual fusion mechanisms, MASTAVN enhances spatial reasoning and temporal synchronization. Through rigorous evaluation in photorealistic 3D simulators (Replica and Matterport3D), MASTAVN achieves significant reductions in task completion time and notable improvements in navigation success rates compared to single-agent and non-collaborative baselines. This highlights the essential role of spatiotemporal coordination in multi-agent systems. Our findings validate MASTAVN's effectiveness in time-sensitive emergency scenarios and establish a paradigm for advancing scalable multi-agent embodied intelligence in complex 3D environments.