Pei Xu

Pei Xu, Jean-Bernard Hayet, Ioannis Karamouzas

Predicting pedestrian movement is critical for human behavior analysis and also for safe and efficient human-agent interactions. However, despite significant advancements, it is still challenging for existing approaches to capture the uncertainty and multimodality of human navigation decision making. In this paper, we propose SocialVAE, a novel approach for human trajectory prediction. The core of SocialVAE is a timewise variational autoencoder architecture that exploits stochastic recurrent neural networks to perform prediction, combined with a social attention mechanism and a backward posterior approximation to allow for better extraction of pedestrian navigation strategies. We show that SocialVAE improves current state-of-the-art performance on several pedestrian trajectory prediction benchmarks, including the ETH/UCY benchmark, Stanford Drone Dataset, and SportVU NBA movement dataset. Code is available at: https://github.com/xupei0610/SocialVAE.

Pei Xu, Kaixiang Xie, Sheldon Andrews et al.

Motivated by humans' ability to adapt skills in the learning of new ones, this paper presents AdaptNet, an approach for modifying the latent space of existing policies to allow new behaviors to be quickly learned from like tasks in comparison to learning from scratch. Building on top of a given reinforcement learning controller, AdaptNet uses a two-tier hierarchy that augments the original state embedding to support modest changes in a behavior and further modifies the policy network layers to make more substantive changes. The technique is shown to be effective for adapting existing physics-based controllers to a wide range of new styles for locomotion, new task targets, changes in character morphology and extensive changes in environment. Furthermore, it exhibits significant increase in learning efficiency, as indicated by greatly reduced training times when compared to training from scratch or using other approaches that modify existing policies. Code is available at https://motion-lab.github.io/AdaptNet.

Pei Xu, Jean-Bernard Hayet, Ioannis Karamouzas

Real-time, accurate prediction of human steering behaviors has wide applications, from developing intelligent traffic systems to deploying autonomous driving systems in both real and simulated worlds. In this paper, we present ContextVAE, a context-aware approach for multi-modal vehicle trajectory prediction. Built upon the backbone architecture of a timewise variational autoencoder, ContextVAE observation encoding employs a dual attention mechanism that accounts for the environmental context and the dynamic agents' states, in a unified way. By utilizing features extracted from semantic maps during agent state encoding, our approach takes into account both the social features exhibited by agents on the scene and the physical environment constraints to generate map-compliant and socially-aware trajectories. We perform extensive testing on the nuScenes prediction challenge, Lyft Level 5 dataset and Waymo Open Motion Dataset to show the effectiveness of our approach and its state-of-the-art performance. In all tested datasets, ContextVAE models are fast to train and provide high-quality multi-modal predictions in real-time. Our code is available at: https://github.com/xupei0610/ContextVAE.

Yanchang Fu, Qiyue Yin, Shengda Liu et al.

Excessive abstraction is a critical challenge in hand abstraction-a task specific to games like Texas hold'em-when solving large-scale imperfect-information games, as it impairs AI performance. This issue arises from extreme implementations of imperfect-recall abstraction, which entirely discard historical information. This paper presents KrwEmd, the first practical algorithm designed to address this problem. We first introduce the k-recall winrate feature, which not only qualitatively distinguishes signal observation infosets by leveraging both future and, crucially, historical game information, but also quantitatively captures their similarity. We then develop the KrwEmd algorithm, which clusters signal observation infosets using earth mover's distance to measure discrepancies between their features. Experimental results demonstrate that KrwEmd significantly improves AI gameplay performance compared to existing algorithms.

Yanchang Fu, Shengda Liu, Pei Xu et al.

High-quality information set abstraction remains a core challenge in solving large-scale imperfect-information extensive-form games (IIEFGs)-such as no-limit Texas Hold'em-where the finite nature of spatial resources hinders strategy solving over the full game. State-of-the-art AI methods rely on pre-trained discrete clustering for abstraction, yet their hard classification irreversibly loses critical information: specifically, the quantifiable subtle differences between information sets-vital for strategy solving-thereby compromising the quality of such solving. Inspired by the word embedding paradigm in natural language processing, this paper proposes the Embedding CFR algorithm, a novel approach for solving strategies in IIEFGs within an embedding space. The algorithm pre-trains and embeds features of isolated information sets into an interconnected low-dimensional continuous space, where the resulting vectors more precisely capture both the distinctions and connections between information sets. Embedding CFR presents a strategy-solving process driven by regret accumulation and strategy updates within this embedding space, with accompanying theoretical analysis verifying its capacity to reduce cumulative regret. Experiments on poker show that with the same spatial overhead, Embedding CFR achieves significantly faster exploitability convergence compared to cluster-based abstraction algorithms, confirming its effectiveness. Furthermore, to our knowledge, it is the first algorithm in poker AI that pre-trains information set abstractions through low-dimensional embedding for strategy solving.

Pei Xu, Ioannis Karamouzas

Data-driven methods for physics-based character control using reinforcement learning have been successfully applied to generate high-quality motions. However, existing approaches typically rely on Gaussian distributions to represent the action policy, which can prematurely commit to suboptimal actions when solving high-dimensional continuous control problems for highly-articulated characters. In this paper, to improve the learning performance of physics-based character controllers, we propose a framework that considers a particle-based action policy as a substitute for Gaussian policies. We exploit particle filtering to dynamically explore and discretize the action space, and track the posterior policy represented as a mixture distribution. The resulting policy can replace the unimodal Gaussian policy which has been the staple for character control problems, without changing the underlying model architecture of the reinforcement learning algorithm used to perform policy optimization. We demonstrate the applicability of our approach on various motion capture imitation tasks. Baselines using our particle-based policies achieve better imitation performance and speed of convergence as compared to corresponding implementations using Gaussians, and are more robust to external perturbations during character control. Related code is available at: https://motion-lab.github.io/PFPN.

Shiyue Cao, Pei Xu, Likun Yang et al.

We study the problem of deceptive path planning (DPP), where an agent aims to conceal its true destination from external observers. While existing work assumes static, non-learning observers, real-world adversaries-such as in critical goods transportation or military operations-can adapt by learning from historical trajectories. To address this gap, we introduce Repeated Deceptive Path Planning (RDPP), a new formulation that explicitly models learnable observers. We show that existing DPP methods fail under this setting, as they cannot adapt to evolving adversarial predictions. While incorporating observer previous predictions into updates enables some adaptation, such incremental updates cause accumulative lag that degrades deception. To this end, we propose Deceptive Meta Planning (DeMP), a two-level optimization framework that combines episode-level adaptation, which enables short-term policy adjustment to counter updated observer, and meta-level updates, which leverage cross-episode feedback to capture how observers update their models and accelerate adaptation in future episodes. In this way, DeMP mitigates the accumulation of adaptation lag, enabling sustained deception against a learning observer. Experiments across environments demonstrate that DeMP significantly outperforms existing approaches in RDPP while maintaining competitive path cost. Our results highlight the importance of modeling repeated interactions with learnable adversaries, providing new insights into deception and privacy in multi-agent systems.

Shiyue Cao, Pei Xu, Likun Yang et al.

Accurately predicting opponents' behavior from interactions is a fundamental capability for large language model (LLM)-based agents in multi-agent and game-theoretic environments. Existing approaches often entangle opponent modeling with prediction, relying on implicit contextual reasoning and limiting adaptability in dynamic interactions. To this end, we propose Structured Opponent Modeling (SOM), a two-stage opponent modeling framework that distinctly separates opponent model construction and opponent prediction. At the construction stage, SOM employs a Structural Causal Model (SCM), a graph-based formalism for representing dependencies among variables, to capture directed links between opponents' observations and actions, yielding an explicit and structured opponent representation. At the prediction stage, the LLM performs structured reasoning along clear pathways derived from the SCM, improving both prediction accuracy and stability. Extensive experiments on diverse multi-agent benchmarks demonstrate that SOM consistently outperforms state-of-the-art LLM-based reasoning baselines, enabling more accurate and adaptable strategic decision-making in complex and dynamic multi-agent interactions.

Pei Xu, Yufei Ye, Shuchun Sun et al.

We present a data-driven approach for physics-based, muscle-driven dexterous control that enables musculoskeletal hands to perform precise piano playing for novel pieces of music outside the reference dataset. Our approach combines high-frequency muscle-level control with low-frequency latent-space coordination in a hierarchical architecture. At the low level, general single-hand policies are trained via reinforcement learning to generate dynamic muscle-tendon activations while tracking trajectories from a large reference motion dataset. The resulting tracking policies are then distilled into variational autoencoder (VAE) models, yielding smooth and structured latent spaces that abstract away low-level muscle dynamics. For the high level, we train piece-specific policies to operate in this latent space, coordinating bimanual motions based on specific goals, denoted by note events extracted from given musical scores, to synthesize performances beyond the reference data. In addition, we present an enhanced musculoskeletal hand model that supports fine control of fingers for accurate low-level motion tracking and diverse high-level motion synthesis. We evaluate the control pipeline of our approach on a diverse piano repertoire spanning multiple musical styles and technical demands. Results demonstrate that our approach can synthesize coordinated bimanual motions with accurate key presses, and achieve the state-of-the-art performance of piano playing in physics-based dexterous control. We also show that our musculoskeletal hand model demonstrates superior biomechanical stability and tracking precision compared to the existing model, and validate that our musculoskeletal hand model and muscle-driven controller can generate physiologically plausible activation patterns that align with human electromyography (EMG) recordings.

Likui Zhang, Tao Tang, Zhihao Zhan et al.

Recent advances in Visual-Language-Action (VLA) models have shown promising potential for robotic manipulation tasks. However, real-world robotic tasks often involve long-horizon, multi-step problem-solving and require generalization for continual skill acquisition, extending beyond single actions or skills. These challenges present significant barriers for existing VLA models, which use monolithic action decoders trained on aggregated data, resulting in poor scalability. To address these challenges, we propose AtomicVLA, a unified planning-and-execution framework that jointly generates task-level plans, atomic skill abstractions, and fine-grained actions. AtomicVLA constructs a scalable atomic skill library through a Skill-Guided Mixture-of-Experts (SG-MoE), where each expert specializes in mastering generic yet precise atomic skills. Furthermore, we introduce a flexible routing encoder that automatically assigns dedicated atomic experts to new skills, enabling continual learning. We validate our approach through extensive experiments. In simulation, AtomicVLA outperforms $π_{0}$ by 2.4\% on LIBERO, 10\% on LIBERO-LONG, and outperforms $π_{0}$ and $π_{0.5}$ by 0.22 and 0.25 in average task length on CALVIN. Additionally, our AtomicVLA consistently surpasses baselines by 18.3\% and 21\% in real-world long-horizon tasks and continual learning. These results highlight the effectiveness of atomic skill abstraction and dynamic expert composition for long-horizon and lifelong robotic tasks. The project page is \href{https://zhanglk9.github.io/atomicvla-web/}{here}.

Yanchang Fu, Qiyue Yin, Shengda Liu et al.

Hand abstraction is crucial for scaling imperfect-information games (IIGs) such as Texas Hold'em, yet progress is limited by the lack of a formal task model and by evaluations that require resource-intensive strategy solving. We introduce signal observation ordered games (SOOGs), a subclass of IIGs tailored to hold'em-style games that cleanly separates signal from player action sequences, providing a precise mathematical foundation for hand abstraction. Within this framework, we define a resolution bound-an information-theoretic upper bound on achievable performance under a given signal abstraction. Using the bound, we show that mainstream outcome-based imperfect-recall algorithms suffer substantial losses by arbitrarily discarding historical information; we formalize this behavior via potential-aware outcome Isomorphism (PAOI) and prove that PAOI characterizes their resolution bound. To overcome this limitation, we propose full-recall outcome isomorphism (FROI), which integrates historical information to raise the bound and improve policy quality. Experiments on hold'em-style benchmarks confirm that FROI consistently outperforms outcome-based imperfect-recall baselines. Our results provide a unified formal treatment of hand abstraction and practical guidance for designing higher-resolution abstractions in IIGs.

Qiyue Yin, Pei Xu, Qiaozhe Li et al.

Recent breakthroughs in Large Language Models (LLMs) have led to a qualitative leap in artificial intelligence' s performance on reasoning tasks, particularly demonstrating remarkable capabilities in mathematical, symbolic, and commonsense reasoning. However, as a critical component of advanced human cognition, strategic reasoning, i.e., the ability to assess multi-agent behaviors in dynamic environments, formulate action plans, and adapt strategies, has yet to be systematically evaluated or modeled. To address this gap, this paper introduces WGSR-Bench, the first strategy reasoning benchmark for LLMs using wargame as its evaluation environment. Wargame, a quintessential high-complexity strategic scenario, integrates environmental uncertainty, adversarial dynamics, and non-unique strategic choices, making it an effective testbed for assessing LLMs' capabilities in multi-agent decision-making, intent inference, and counterfactual reasoning. WGSR-Bench designs test samples around three core tasks, i.e., Environmental situation awareness, Opponent risk modeling and Policy generation, which serve as the core S-POE architecture, to systematically assess main abilities of strategic reasoning. Finally, an LLM-based wargame agent is designed to integrate these parts for a comprehensive strategy reasoning assessment. With WGSR-Bench, we hope to assess the strengths and limitations of state-of-the-art LLMs in game-theoretic strategic reasoning and to advance research in large model-driven strategic intelligence.

Jianhua Han, Meng Tian, Jiangtong Zhu et al.

Autonomous driving heavily relies on accurate and robust spatial perception. Many failures arise from inaccuracies and instability, especially in long-tail scenarios and complex interactions. However, current vision-language models are weak at spatial grounding and understanding, and VLA systems built on them therefore show limited perception and localization ability. To address these challenges, we introduce Percept-WAM, a perception-enhanced World-Awareness-Action Model that is the first to implicitly integrate 2D/3D scene understanding abilities within a single vision-language model (VLM). Instead of relying on QA-style spatial reasoning, Percept-WAM unifies 2D/3D perception tasks into World-PV and World-BEV tokens, which encode both spatial coordinates and confidence. We propose a grid-conditioned prediction mechanism for dense object perception, incorporating IoU-aware scoring and parallel autoregressive decoding, improving stability in long-tail, far-range, and small-object scenarios. Additionally, Percept-WAM leverages pretrained VLM parameters to retain general intelligence (e.g., logical reasoning) and can output perception results and trajectory control outputs directly. Experiments show that Percept-WAM matches or surpasses classical detectors and segmenters on downstream perception benchmarks, achieving 51.7/58.9 mAP on COCO 2D detection and nuScenes BEV 3D detection. When integrated with trajectory decoders, it further improves planning performance on nuScenes and NAVSIM, e.g., surpassing DiffusionDrive by 2.1 in PMDS on NAVSIM. Qualitative results further highlight its strong open-vocabulary and long-tail generalization.

Pei Xu, Zhen Wu, Ruocheng Wang et al.

Learning a control policy for a multi-phase, long-horizon task, such as basketball maneuvers, remains challenging for reinforcement learning approaches due to the need for seamless policy composition and transitions between skills. A long-horizon task typically consists of distinct subtasks with well-defined goals, separated by transitional subtasks with unclear goals but critical to the success of the entire task. Existing methods like the mixture of experts and skill chaining struggle with tasks where individual policies do not share significant commonly explored states or lack well-defined initial and terminal states between different phases. In this paper, we introduce a novel policy integration framework to enable the composition of drastically different motor skills in multi-phase long-horizon tasks with ill-defined intermediate states. Based on that, we further introduce a high-level soft router to enable seamless and robust transitions between the subtasks. We evaluate our framework on a set of fundamental basketball skills and challenging transitions. Policies trained by our approach can effectively control the simulated character to interact with the ball and accomplish the long-horizon task specified by real-time user commands, without relying on ball trajectory references.

Yuxiang Mai, Qiyue Yin, Wancheng Ni et al.

In recent years, diversity has emerged as a useful mechanism to enhance the efficiency of multi-agent reinforcement learning (MARL). However, existing methods predominantly focus on designing policies based on individual agent characteristics, often neglecting the interplay and mutual influence among agents during policy formation. To address this gap, we propose Competitive Diversity through Constructive Conflict (CoDiCon), a novel approach that incorporates competitive incentives into cooperative scenarios to encourage policy exchange and foster strategic diversity among agents. Drawing inspiration from sociological research, which highlights the benefits of moderate competition and constructive conflict in group decision-making, we design an intrinsic reward mechanism using ranking features to introduce competitive motivations. A centralized intrinsic reward module generates and distributes varying reward values to agents, ensuring an effective balance between competition and cooperation. By optimizing the parameterized centralized reward module to maximize environmental rewards, we reformulate the constrained bilevel optimization problem to align with the original task objectives. We evaluate our algorithm against state-of-the-art methods in the SMAC and GRF environments. Experimental results demonstrate that CoDiCon achieves superior performance, with competitive intrinsic rewards effectively promoting diverse and adaptive strategies among cooperative agents.

Zhen Wu, Jiaman Li, Pei Xu et al.

Intelligent agents must autonomously interact with the environments to perform daily tasks based on human-level instructions. They need a foundational understanding of the world to accurately interpret these instructions, along with precise low-level movement and interaction skills to execute the derived actions. In this work, we propose the first complete system for synthesizing physically plausible, long-horizon human-object interactions for object manipulation in contextual environments, driven by human-level instructions. We leverage large language models (LLMs) to interpret the input instructions into detailed execution plans. Unlike prior work, our system is capable of generating detailed finger-object interactions, in seamless coordination with full-body movements. We also train a policy to track generated motions in physics simulation via reinforcement learning (RL) to ensure physical plausibility of the motion. Our experiments demonstrate the effectiveness of our system in synthesizing realistic interactions with diverse objects in complex environments, highlighting its potential for real-world applications.

Pei Xu, Xiumin Shang, Victor Zordan et al.

We present a deep learning method for composite and task-driven motion control for physically simulated characters. In contrast to existing data-driven approaches using reinforcement learning that imitate full-body motions, we learn decoupled motions for specific body parts from multiple reference motions simultaneously and directly by leveraging the use of multiple discriminators in a GAN-like setup. In this process, there is no need of any manual work to produce composite reference motions for learning. Instead, the control policy explores by itself how the composite motions can be combined automatically. We further account for multiple task-specific rewards and train a single, multi-objective control policy. To this end, we propose a novel framework for multi-objective learning that adaptively balances the learning of disparate motions from multiple sources and multiple goal-directed control objectives. In addition, as composite motions are typically augmentations of simpler behaviors, we introduce a sample-efficient method for training composite control policies in an incremental manner, where we reuse a pre-trained policy as the meta policy and train a cooperative policy that adapts the meta one for new composite tasks. We show the applicability of our approach on a variety of challenging multi-objective tasks involving both composite motion imitation and multiple goal-directed control.

Pei Xu, Ioannis Karamouzas

We present a simple and intuitive approach for interactive control of physically simulated characters. Our work builds upon generative adversarial networks (GAN) and reinforcement learning, and introduces an imitation learning framework where an ensemble of classifiers and an imitation policy are trained in tandem given pre-processed reference clips. The classifiers are trained to discriminate the reference motion from the motion generated by the imitation policy, while the policy is rewarded for fooling the discriminators. Using our GAN-based approach, multiple motor control policies can be trained separately to imitate different behaviors. In runtime, our system can respond to external control signal provided by the user and interactively switch between different policies. Compared to existing methods, our proposed approach has the following attractive properties: 1) achieves state-of-the-art imitation performance without manually designing and fine tuning a reward function; 2) directly controls the character without having to track any target reference pose explicitly or implicitly through a phase state; and 3) supports interactive policy switching without requiring any motion generation or motion matching mechanism. We highlight the applicability of our approach in a range of imitation and interactive control tasks, while also demonstrating its ability to withstand external perturbations as well as to recover balance. Overall, our approach generates high-fidelity motion, has low runtime cost, and can be easily integrated into interactive applications and games.

Pei Xu, Ioannis Karamouzas

Despite significant advancements in the field of multi-agent navigation, agents still lack the sophistication and intelligence that humans exhibit in multi-agent settings. In this paper, we propose a framework for learning a human-like general collision avoidance policy for agent-agent interactions in fully decentralized, multi-agent environments. Our approach uses knowledge distillation with reinforcement learning to shape the reward function based on expert policies extracted from human trajectory demonstrations through behavior cloning. We show that agents trained with our approach can take human-like trajectories in collision avoidance and goal-directed steering tasks not provided by the demonstrations, outperforming the experts as well as learning-based agents trained without knowledge distillation.

Guangda Chen, Lifan Pan, Yu'an Chen et al.

This paper proposes an end-to-end deep reinforcement learning approach for mobile robot navigation with dynamic obstacles avoidance. Using experience collected in a simulation environment, a convolutional neural network (CNN) is trained to predict proper steering actions of a robot from its egocentric local occupancy maps, which accommodate various sensors and fusion algorithms. The trained neural network is then transferred and executed on a real-world mobile robot to guide its local path planning. The new approach is evaluated both qualitatively and quantitatively in simulation and real-world robot experiments. The results show that the map-based end-to-end navigation model is easy to be deployed to a robotic platform, robust to sensor noise and outperforms other existing DRL-based models in many indicators.

Pei Xu, Shan Huang, Hongzhen Wang et al.

Chinese keyword spotting is a challenging task as there is no visual blank for Chinese words. Different from English words which are split naturally by visual blanks, Chinese words are generally split only by semantic information. In this paper, we propose a new Chinese keyword spotter for natural images, which is inspired by Mask R-CNN. We propose to predict the keyword masks guided by text line detection. Firstly, proposals of text lines are generated by Faster R-CNN;Then, text line masks and keyword masks are predicted by segmentation in the proposals. In this way, the text lines and keywords are predicted in parallel. We create two Chinese keyword datasets based on RCTW-17 and ICPR MTWI2018 to verify the effectiveness of our method.

Pei Xu

The uncertainty and variability of underwater environment propose the request to control underwater robots in real time and dynamically, especially in the scenarios where human and robots need to work collaboratively in the field. However, the underwater environment imposes harsh restrictions on the application of typical control and communication methods. Considering that gestures are a natural and efficient interactive way for human, we, utilizing convolution neural network, implement a real-time gesture-based recognition system, who can recognize 50 kinds of gestures from images captured by one normal monocular camera, and apply this recognition system in human and underwater robot interaction. We design A Flexible and Extendable Interaction Scheme (AFEIS) through which underwater robots can be programmed in situ underwater by human operators using customized gesture-based sign language. This paper elaborates the design of gesture recognition system and AFEIS, and presents our field trial results when applying this system and scheme on underwater robots.

Baoguang Shi, Cong Yao, Minghui Liao et al.

Chinese is the most widely used language in the world. Algorithms that read Chinese text in natural images facilitate applications of various kinds. Despite the large potential value, datasets and competitions in the past primarily focus on English, which bares very different characteristics than Chinese. This report introduces RCTW, a new competition that focuses on Chinese text reading. The competition features a large-scale dataset with 12,263 annotated images. Two tasks, namely text localization and end-to-end recognition, are set up. The competition took place from January 20 to May 31, 2017. 23 valid submissions were received from 19 teams. This report includes dataset description, task definitions, evaluation protocols, and results summaries and analysis. Through this competition, we call for more future research on the Chinese text reading problem. The official website for the competition is http://rctw.vlrlab.net

Pei Xu

In this project, we design a real-time human-computer interaction system based on hand gesture. The whole system consists of three components: hand detection, gesture recognition and human-computer interaction (HCI) based on recognition; and realizes the robust control of mouse and keyboard events with a higher accuracy of gesture recognition. Specifically, we use the convolutional neural network (CNN) to recognize gestures and makes it attainable to identify relatively complex gestures using only one cheap monocular camera. We introduce the Kalman filter to estimate the hand position based on which the mouse cursor control is realized in a stable and smooth way. During the HCI stage, we develop a simple strategy to avoid the false recognition caused by noises - mostly transient, false gestures, and thus to improve the reliability of interaction. The developed system is highly extendable and can be used in human-robotic or other human-machine interaction scenarios with more complex command formats rather than just mouse and keyboard events.