Searcharxiv⌕ Search

arXiv subjects

Yuhong Cao

Publications and source records attributed to Yuhong Cao.

At least 19 recordsLinked to original sources
arXiv2026

PRIMAL3: Pathfinding via Reinforcement and Imitation Multi-Agent Learning - Leveraging LaCAM3

We present PRIMAL3, an ultra-large-scale learning-based framework for multi-agent pathfinding (MAPF) that integrates reinforcement learning, topology-aware communication, LaCAM3-guided training, and PIBT-based action refinement. PRIMAL3 targets failures at topologically critical states, where agents must coordinate decisively around bottlenecks, dead ends, and persistent conflicts. Each agent is represented using features derived from cut vertices, dead-end regions, shortest-path distances, and blocking estimates. Two complementary graphs capture agent interactions: a same-direction following graph propagates multihop context along compatible paths, while a different-direction conflict graph differentiates agents competing for shared space through masked attention and relative features. During training, we propose to let policy entropy identify uncertain agents, for which LaCAM3 provides confidence-triggered action interventions and label-smoothed imitation targets. During execution, a priority-aware PIBT module refines the proposed joint actions using persistent, learned, and distance-aware priorities together with policy-aware fallback preferences while maintaining collision-free execution. The resulting framework combines learned exploration with structured expert guidance without requiring LaCAM3 at inference. Experiments demonstrate that PRIMAL3 substantially outperforms state-of-the-art learning-based baselines and scales to ultra-large instances with up to city-level 100,000 agents. Real-world experiments further demonstrate the feasibility of deploying PRIMAL3 on physical robotic systems and ablation studies validate the individual contributions the components we proposed. Project page: https://marmotlab.github.io/PRIMAL3/

cs.RO↗
arXiv2026

KiVi: Kinesthetic-Visuospatial Integration for Dynamic and Safe Egocentric Legged Locomotion

Vision-based locomotion has shown great promise in enabling legged robots to perceive and adapt to complex environments. However, visual information is inherently fragile, being vulnerable to occlusions, reflections, and lighting changes, which often cause instability in locomotion. Inspired by animal sensorimotor integration, we propose KiVi, a Kinesthetic-Visuospatial integration framework, where kinesthetics encodes proprioceptive sensing of body motion and visuospatial reasoning captures visual perception of surrounding terrain. Specifically, KiVi separates these pathways, leveraging proprioception as a stable backbone while selectively incorporating vision for terrain awareness and obstacle avoidance. This modality-balanced, yet integrative design, combined with memory-enhanced attention, allows the robot to robustly interpret visual cues while maintaining fallback stability through proprioception. Extensive experiments show that our method enables quadruped robots to stably traverse diverse terrains and operate reliably in unstructured outdoor environments, remaining robust to out-of-distribution(OOD) visual noise and occlusion unseen during training, thereby highlighting its effectiveness and applicability to real-world legged locomotion. Project Page: https://marmotlab.github.io/kivi-quadruped/

cs.RO↗
arXiv2026

APEX: Action Priors Enable Efficient Exploration for Robust Motion Tracking on Legged Robots

Learning natural, animal-like locomotion from demonstrations has become a core paradigm in legged robotics. While motion tracking can reproduce reference gaits, many approaches still require substantial tuning and depend on reference motion inputs at deployment, which can limit responsiveness to task objectives and reduce adaptability. We present APEX (Action Priors enable Efficient eXploration), a motion-tracking reinforcement learning (RL) framework that removes deployment-time dependence on reference motion inputs, improves sample efficiency, and reduces tuning effort. APEX integrates demonstrations into RL via decaying action priors, which guide early exploration toward demonstration-consistent actions and then fade to zero, yielding a pure RL policy at deployment. This is combined with a multi-critic framework that separates style and task + regularization learning signals. Moreover, APEX enables a single policy to learn diverse motions and transfer reference-like styles across different terrains and velocities, while remaining robust to variations in training parameters. We validate our method in simulation on both humanoid and quadruped robots, and with zero-shot deployment on a Unitree Go2 robot. Website and code: https://marmotlab.github.io/APEX/.

cs.RO↗
arXiv2026

Compositional gradient engineering for enhanced ferroelectricity in ultrathin AlScN

Ferroelectric AlScN is promising for CMOS-compatible non-volatile memory, but thickness scaling is limited by leakage, premature breakdown, and defect-mediated failure. Here we show that compositional grading within a continuous wurtzite AlN-AlScN lattice mitigates these limitations by distributing structural and polarization discontinuities across the film thickness, reducing defect formation and local field concentration. In a 20 nm graded heterostructure, monotonic Sc incorporation and AlN-rich boundaries produce reversible ferroelectric switching, an as-grown metal-polar state, a 21% higher breakdown field, 10% enhanced remanent polarization, and 40x higher resistivity relative to homogeneous AlScN. Time-domain PUND measurements reveal strongly suppressed post-switching leakage, consistent with reduced defect-assisted and polarization-coupled conduction. This improved dielectric robustness enables ferroelectric functionality in 5 nm graded stacks containing only a 2 nm $\mathrm{Al}_{0.64}\mathrm{Sc}_{0.36}\mathrm{N}$ region, with measurable switching near 1 V. These results establish compositional grading as a defect- and field-management strategy for scalable ultrathin wurtzite ferroelectrics.

cond-mat.mtrl-sci↗
arXiv2026

TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion

Agile humanoid locomotion across diverse challenging terrain demands both wide perceptual coverage and precise local geometry understanding. Motivated by the way humans selectively look at relevant terrain during locomotion, we introduce TAGA, a Terrain-aware Active Gaze learning framework for Attention-based humanoid control. By fusing vision, proprioception, and motion commands, our framework guides the model to learn anticipatory cues and actively attend to specific areas of the height scan, selectively using these informative regions for the downstream network. This adaptively increases the information density of observations under tight onboard computational constraints, thus enabling fine-grained perceptive locomotion over larger-scale terrains. We find that such gaze behaviors can naturally emerge through reinforcement learning alone, without requiring additional supervision or explicit guidance, significantly improve training efficiency. As a result, the trained policy demonstrates robust and generalizable locomotion in simulation and on hardware, including reliable terrain-aware foothold selection, elevated-platform traversal, competitive sparse-foothold traversal, and the largest reported real-world gap traversal distance of 1.2m among perceptive humanoid locomotion systems, while maintaining stability under severe perceptual disturbances and environmental interference.

cs.RO↗
arXiv2026

HeLoM: Hierarchical Learning for Whole-Body Loco-Manipulation by a Hexapod Robot

In nature, animals often need to move/manipulate objects comparable in weight/size to their own bodies. Compared to grasping and carrying, pushing provides a more straightforward and efficient non-prehensile manipulation strategy, avoiding complex grasp design while leveraging direct contact to regulate an object's pose during interaction. Achieving effective pushing, however, requires both sufficient manipulation capability and stable whole-body coordination, which is particularly challenging when dealing with heavy or irregular objects. To address these challenges, we propose HeLoM, a learning-based hierarchical whole-body manipulation framework for hexapod robots that exploits coordinated multi-limb control and is applicable to multi-legged robotic systems. Inspired by the cooperative strategies of multi-legged insects, our framework leverages multiple contact points and high degrees of freedom to enable efficient and dynamic whole-body coordination during object interaction. HeLoM's high-level planner plans pushing behaviors, while its low-level controller maintains locomotion stability and generates dynamically consistent joint actions. This design enables the robot to maintain balance while executing continuous and controllable pushing behaviors through coordinated foreleg interaction and supportive hind-leg propulsion. We validate the effectiveness of HeLoM through both simulation and real-world experiments. Results show that our framework can stably push objects of varying sizes and unknown physical properties to designated goal poses in the real world.

cs.RO↗
arXiv2026

ORION: Option-Regularized Deep Reinforcement Learning for Cooperative Multi-Agent Online Navigation

Existing methods for multi-agent navigation typically assume fully known environments, offering limited support for partially known scenarios with outdated or imperfect prior maps, such as warehouses or factory floors. There, agents need to balance path optimality with collecting and sharing environmental information to help teammates reach their own targets. To these ends, we propose ORION, a novel deep reinforcement learning framework for cooperative multi-agent online navigation in partially known environments. Starting from an imperfect prior map, ORION trains agents to make decentralized decisions, coordinate toward individual targets, and actively reduce task-relevant map uncertainty through online observation sharing in a closed perception-action loop. We first design a shared graph encoder that fuses prior map with online perception into a unified representation, providing robust state embeddings under environmental discrepancies. At the core of ORION is an option-critic framework that learns high-level cooperative modes translated into sequences of low-level actions, enabling adaptive switching between individual navigation and team-level exploration. We further introduce a dual-stage cooperation strategy that allows agents to assist teammates under map uncertainty, thereby reducing the overall makespan. Across extensive maze-like maps and large-scale warehouse environments, ORION achieves high-quality real-time decentralized cooperation while scaling to up to 10 robots, outperforming state-of-the-art classical and learning-based baselines. Finally, we validate ORION on physical robot teams, demonstrating its robustness and practicality for real-world cooperative navigation.

cs.RO↗
arXiv2026

AID: Agent Intent from Diffusion for Multi-Agent Informative Path Planning

Information gathering in large-scale or time-critical scenarios (e.g., environmental monitoring, search and rescue) requires broad coverage within limited time budgets, motivating the use of multi-agent systems. These scenarios are commonly formulated as multi-agent informative path planning (MAIPP), where multiple agents must coordinate to maximize information gain while operating under budget constraints. A central challenge in MAIPP is ensuring effective coordination while the belief over the environment evolves with incoming measurements. Recent learning-based approaches address this by using distributions over future positions as "intent" to support coordination. However, these autoregressive intent predictors are computationally expensive and prone to compounding errors. Inspired by the effectiveness of diffusion models as expressive, long-horizon policies, we propose AID, a fully decentralized MAIPP framework that leverages diffusion models to generate long-term trajectories in a non-autoregressive manner. AID first performs behavior cloning on trajectories produced by existing MAIPP planners and then fine-tunes the policy using reinforcement learning via Diffusion Policy Policy Optimization (DPPO). This two-stage pipeline enables the policy to inherit expert behavior while learning improved coordination through online reward feedback. Experiments demonstrate that AID consistently improves upon the MAIPP planners it is trained from, achieving 4x faster execution and up to 17% increased information gain, while scaling effectively to larger numbers of agents. Our implementation is publicly available at https://github.com/marmotlab/AID.

cs.RO↗
arXiv2026

COIN: Collaborative Interaction-Aware Multi-Agent Reinforcement Learning for Self-Driving Systems

Multi-Agent Self-Driving (MASD) systems provide an effective solution for coordinating autonomous vehicles to reduce congestion and enhance both safety and operational efficiency in future intelligent transportation systems. Multi-Agent Reinforcement Learning (MARL) has emerged as a promising approach for developing advanced end-to-end MASD systems. However, achieving efficient and safe collaboration in dynamic MASD systems remains a significant challenge in dense scenarios with complex agent interactions. To address this challenge, we propose a novel collaborative(CO-) interaction-aware(-IN) MARL framework, named COIN. Specifically, we develop a new counterfactual individual-global twin delayed deep deterministic policy gradient (CIG-TD3) algorithm, crafted in a "centralized training, decentralized execution" (CTDE) manner, which aims to jointly optimize the individual objectives (navigation) and the global objectives (collaboration) of agents. We further introduce a dual-level interaction-aware centralized critic architecture that captures both local pairwise interactions and global system-level dependencies, enabling more accurate global value estimation and improved credit assignment for collaborative policy learning. We conduct extensive simulation experiments in dense urban traffic environments, which demonstrate that COIN consistently outperforms other advanced baseline methods in both safety and efficiency across various system sizes. These results highlight its superiority in complex and dynamic MASD scenarios, as further validated through real-world robot demonstrations. Supplementary videos are available at https://marmotlab.github.io/COIN/

cs.RO↗
arXiv2026

Multimodal Fused Learning for Solving the Generalized Traveling Salesman Problem in Robotic Task Planning

Effective and efficient task planning is essential for mobile robots, especially in applications like warehouse retrieval and environmental monitoring. These tasks often involve selecting one location from each of several target clusters, forming a Generalized Traveling Salesman Problem (GTSP) that remains challenging to solve both accurately and efficiently. To address this, we propose a Multimodal Fused Learning (MMFL) framework that leverages both graph and image-based representations to capture complementary aspects of the problem, and learns a policy capable of generating high-quality task planning schemes in real time. Specifically, we first introduce a coordinate-based image builder that transforms GTSP instances into spatially informative representations. We then design an adaptive resolution scaling strategy to enhance adaptability across different problem scales, and develop a multimodal fusion module with dedicated bottlenecks that enables effective integration of geometric and spatial features. Extensive experiments show that our MMFL approach significantly outperforms state-of-the-art methods across various GTSP instances while maintaining the computational efficiency required for real-time robotic applications. Physical robot tests further validate its practical effectiveness in real-world scenarios.

cs.AI↗
arXiv2026

ImagiNav: Scalable Embodied Navigation via Generative Visual Prediction and Inverse Dynamics

Enabling robots to navigate open-world environments via natural language is critical for general-purpose autonomy. Yet, Vision-Language Navigation has relied on end-to-end policies trained on expensive, embodiment-specific robot data. While recent foundation models trained on vast simulation data show promise, the challenge of scaling and generalizing due to the limited scene diversity and visual fidelity in simulation persists. To address this gap, we propose ImagiNav, a novel modular paradigm that decouples visual planning from robot actuation, enabling the direct utilization of diverse in-the-wild navigation videos. Our framework operates as a hierarchy: a Vision-Language Model first decomposes instructions into textual subgoals; a finetuned generative video model then imagines the future video trajectory towards that subgoal; finally, an inverse dynamics model extracts the trajectory from the imagined video, which can then be tracked via a low-level controller. We additionally develop a scalable data pipeline of in-the-wild navigation videos auto-labeled via inverse dynamics and a pretrained Vision-Language Model. ImagiNav demonstrates strong zero-shot transfer to robot navigation without requiring robot demonstrations, paving the way for generalist robots that learn navigation directly from unlabeled, open-world data.

cs.RO↗
arXiv2026

Controlled non-volatile modulation of optical dispersion in monolayer tungsten disulfide via ferroelectric polarization patterning

The manipulation of optical properties, including reflection, refraction, polarization, phase, and frequency, has long been central to advancing photonic and optoelectronic technologies. However, existing electro-optical approaches rely on volatile mechanisms that require continuous power consumption. Here, we demonstrate strong, nonvolatile modulation of optical dispersion in monolayer tungsten disulfide (ML WS2) using patterned ferroelectric domains in aluminum scandium nitride (AlScN). By locally poling ferroelectric domains into opposite states, we achieve substantial manipulation of the complex refractive index (Delta n > 0.7, Delta k > 0.4) and excitonic energy shifts (~50 meV) in ML WS2, comparable to previous gate-tuning approaches while eliminating continuous power consumption. We introduce an asymmetric screening model that reveals how ferroelectric polarization induces carrier-density-dependent Coulomb screening, leading to distinct excitonic behaviors between electron- and hole-doped regions. Furthermore, we demonstrate a gate-free lateral p-n homojunction with a rectification ratio of 6e10^3, formed through spatial carrier redistribution. These findings establish ferroelectric/2D heterostructures as a powerful platform for nonvolatile optical dispersion engineering, enabling energy-efficient, reconfigurable photonic and optoelectronic devices.

physics.optics↗
arXiv2026

GPO: Growing Policy Optimization for Legged Robot Locomotion and Whole-Body Control

Training reinforcement learning (RL) policies for legged robots remains challenging due to high-dimensional continuous actions, hardware constraints, and limited exploration. Existing methods for locomotion and whole-body control work well for position-based control with environment-specific heuristics (e.g., reward shaping, curriculum design, and manual initialization), but are less effective for torque-based control, where sufficiently exploring the action space and obtaining informative gradient signals for training is significantly more difficult. We introduce Growing Policy Optimization (GPO), a training framework that applies a time-varying action transformation to restrict the effective action space in the early stage, thereby encouraging more effective data collection and policy learning, and then progressively expands it to enhance exploration and achieve higher expected return. We prove that this transformation preserves the PPO update rule and introduces only bounded, vanishing gradient distortion, thereby ensuring stable training. We evaluate GPO on both quadruped and hexapod robots, including zero-shot deployment of simulation-trained policies on hardware. Policies trained with GPO consistently achieve better performance. These results suggest that GPO provides a general, environment-agnostic optimization framework for learning legged locomotion.

cs.RO↗
arXiv2026

FARE: Fast-Slow Agentic Robotic Exploration

This work advances autonomous robot exploration by integrating agent-level semantic reasoning with fast local control. We introduce FARE, a hierarchical autonomous exploration framework that integrates a large language model (LLM) for global reasoning with a reinforcement learning (RL) policy for local decision making. FARE follows a fast-slow thinking paradigm. The slow-thinking LLM module interprets a concise textual description of the unknown environment and synthesizes an agent-level exploration strategy, which is then grounded into a sequence of global waypoints through a topological graph. To further improve reasoning efficiency, this module employs a modularity-based pruning mechanism that reduces redundant graph structures. The fast-thinking RL module executes exploration by reacting to local observations while being guided by the LLM-generated global waypoints. The RL policy is additionally shaped by a reward term that encourages adherence to the global waypoints, enabling coherent and robust closed-loop behavior. This architecture decouples semantic reasoning from geometric decision, allowing each module to operate in its appropriate temporal and spatial scale. In challenging simulated environments, our results show that FARE achieves substantial improvements in exploration efficiency over state-of-the-art baselines. We further deploy FARE on hardware and validate it in complex, large scale $200m\times130m$ building environment.

cs.RO↗
arXiv2025

APEX: Action Priors Enable Efficient Exploration for Robust Motion Tracking on Legged Robots

Learning natural, animal-like locomotion from demonstrations has become a core paradigm in legged robotics. Despite the recent advancements in motion tracking, most existing methods demand extensive tuning and rely on reference data during deployment, limiting adaptability. We present APEX (Action Priors enable Efficient Exploration), a plug-and-play extension to state-of-the-art motion tracking algorithms that eliminates any dependence on reference data during deployment, improves sample efficiency, and reduces parameter tuning effort. APEX integrates expert demonstrations directly into reinforcement learning (RL) by incorporating decaying action priors, which initially bias exploration toward expert demonstrations but gradually allow the policy to explore independently. This is combined with a multi-critic framework that balances task performance with motion style. Moreover, APEX enables a single policy to learn diverse motions and transfer reference-like styles across different terrains and velocities, while remaining robust to variations in reward design. We validate the effectiveness of our method through extensive experiments in both simulation and on a Unitree Go2 robot. By leveraging demonstrations to guide exploration during RL training, without imposing explicit bias toward them, APEX enables legged robots to learn with greater stability, efficiency, and generalization. We believe this approach paves the way for guidance-driven RL to boost natural skill acquisition in a wide array of robotic tasks, from locomotion to manipulation. Website and code: https://marmotlab.github.io/APEX/.

cs.RO↗
arXiv2025

Search-TTA: A Multimodal Test-Time Adaptation Framework for Visual Search in the Wild

To perform outdoor visual navigation and search, a robot may leverage satellite imagery to generate visual priors. This can help inform high-level search strategies, even when such images lack sufficient resolution for target recognition. However, many existing informative path planning or search-based approaches either assume no prior information, or use priors without accounting for how they were obtained. Recent work instead utilizes large Vision Language Models (VLMs) for generalizable priors, but their outputs can be inaccurate due to hallucination, leading to inefficient search. To address these challenges, we introduce Search-TTA, a multimodal test-time adaptation framework with a flexible plug-and-play interface compatible with various input modalities (e.g., image, text, sound) and planning methods (e.g., RL-based). First, we pretrain a satellite image encoder to align with CLIP's visual encoder to output probability distributions of target presence used for visual search. Second, our TTA framework dynamically refines CLIP's predictions during search using uncertainty-weighted gradient updates inspired by Spatial Poisson Point Processes. To train and evaluate Search-TTA, we curate AVS-Bench, a visual search dataset based on internet-scale ecological data containing 380k images and taxonomy data. We find that Search-TTA improves planner performance by up to 30.0%, particularly in cases with poor initial CLIP predictions due to domain mismatch and limited training data. It also performs comparably with significantly larger VLMs, and achieves zero-shot generalization via emergent alignment to unseen modalities. Finally, we deploy Search-TTA on a real UAV via hardware-in-the-loop testing, by simulating its operation within a large-scale simulation that provides onboard sensing.

cs.RO↗
arXiv2025

IR2: Implicit Rendezvous for Robotic Exploration Teams under Sparse Intermittent Connectivity

Information sharing is critical in time-sensitive and realistic multi-robot exploration, especially for smaller robotic teams in large-scale environments where connectivity may be sparse and intermittent. Existing methods often overlook such communication constraints by assuming unrealistic global connectivity. Other works account for communication constraints (by maintaining close proximity or line of sight during information exchange), but are often inefficient. For instance, preplanned rendezvous approaches typically involve unnecessary detours resulting from poorly timed rendezvous, while pursuit-based approaches often result in short-sighted decisions due to their greedy nature. We present IR2, a deep reinforcement learning approach to information sharing for multi-robot exploration. Leveraging attention-based neural networks trained via reinforcement and curriculum learning, IR2 allows robots to effectively reason about the longer-term trade-offs between disconnecting for solo exploration and reconnecting for information sharing. In addition, we propose a hierarchical graph formulation to maintain a sparse yet informative graph, enabling our approach to scale to large-scale environments. We present simulation results in three large-scale Gazebo environments, which show that our approach yields 6.6-34.1% shorter exploration paths when compared to state-of-the-art baselines, and lastly deploy our learned policy on hardware. Our simulation training and testing code is available at https://ir2-explore.github.io.

cs.RO↗
arXiv2025

HEADER: Hierarchical Robot Exploration via Attention-Based Deep Reinforcement Learning with Expert-Guided Reward

This work pushes the boundaries of learning-based methods in autonomous robot exploration in terms of environmental scale and exploration efficiency. We present HEADER, an attention-based reinforcement learning approach with hierarchical graphs for efficient exploration in large-scale environments. HEADER follows existing conventional methods to construct hierarchical representations for the robot belief/map, but further designs a novel community-based algorithm to construct and update a global graph, which remains fully incremental, shape-adaptive, and operates with linear complexity. Building upon attention-based networks, our planner finely reasons about the nearby belief within the local range while coarsely leveraging distant information at the global scale, enabling next-best-viewpoint decisions that consider multi-scale spatial dependencies. Beyond novel map representation, we introduce a parameter-free privileged reward that significantly improves model performance and produces near-optimal exploration behaviors, by avoiding training objective bias caused by handcrafted reward shaping. In simulated challenging, large-scale exploration scenarios, HEADER demonstrates better scalability than most existing learning and non-learning methods, while achieving a significant improvement in exploration efficiency (up to 20%) over state-of-the-art baselines. We also deploy HEADER on hardware and validate it in complex, large-scale real-life scenarios, including a 300m*230m campus environment.

cs.RO↗