Semantically Structured Mixture-of-Experts for Compositional Robotic Manipulation

Diffusion-based policies have established a new standard for precise robotic manipulation but face a critical scalability bottleneck: high-performance models are computationally expensive, while lightweight alternatives often fail to generalize across diverse multi-task environments. Mixture-of-Experts (MoE) architectures offer a promising path to efficiency by activating only a subset of parameters. However, existing MoE routing mechanisms typically rely on low-level noise or latent statistics, ignoring the compositional nature of manipulation tasks. This can fragment reusable behaviors across experts, limiting interpretability and transferability. We introduce Semantically Structured Mixture-of-Experts Diffusion Policy (SMoDP) for compositional robotic manipulation, a framework that grounds expert specialization in semantic task structure. SMoDP leverages a lightweight, inference-time skill predictor, supervised by offline annotations from Vision-Language Models (VLMs), to route action chunks to experts specialized for specific behavioral phases. To ensure robust assignment, we propose a dual contrastive alignment strategy that grounds multi-modal observations in language-defined skill semantics (Inter-modal) while enforcing routing consistency across visually distinct but functionally related behaviors (Intra-modal). Our approach outperforms representative diffusion and MoE-based baselines on multi-task benchmarks with significantly improved parameter efficiency and demonstrates effective compositional transfer to novel tasks through parameter-efficient fine-tuning. Project website: https://deng-cy20.github.io/SMoDP/

Beyond Pixels: Learning Invariant Rewards for Real-World Robotics From a Few Demonstrations

Designing reward functions that generalize beyond controlled laboratory settings remains a fundamental challenge in reinforcement learning for robotics. In open-world manipulation problems, a single task can appear in numerous variants through different object instances, positions, and camera viewpoints. Recent vision-based reward models tend to memorize specific pixel distributions and fail to generalize beyond their training conditions. To address this, we propose a framework that learns invariant symbolic reward functions from as few as five demonstrations. The insight is to shift from visual feature-fitting to the discovery of behavioral invariants: task-level properties that remain constant across diverse visual instantiations. The framework has two coupled components: a structural reward formulation that encodes task-level strategies and physical constraints while preserving optimal policy invariance, and a hybrid symbolic-numerical procedure that distills these invariants from demonstrations without online interaction. Experiments on eight Meta-World tasks and three Franka manipulation tasks demonstrate that our method achieves stronger process alignment and policy rollout ranking abilities compared to baselines, accelerating downstream policy learning. Three real-world out-of-distribution experiments further show that the same learned reward generalizes zero-shot to position, viewpoint, and object variations, enabling a single reward representation to be reused across diverse task variants in practice.

Beyond Action Residuals: Real-World Robot Policy Steering via Bottleneck Latent Reinforcement Learning

Pretrained imitation policies have become a strong foundation for robot manipulation, but they often require online improvement to overcome execution errors, limited dataset coverage, and deployment mismatch. A central question is therefore how reinforcement learning (RL) should adapt policies after offline pretraining. Existing lightweight methods commonly apply residual corrections directly in action space, but this often leads to noisy and poorly structured exploration. In this work, we propose Z-Perturbation Reinforcement Learning (ZPRL), an approach that steers pretrained policies through a compact bottleneck latent rather than through policy weights or output actions. During offline training, we augment the policy with a plug-and-play variational information bottleneck (VIB) module to extract a task-relevant latent interface from observation embeddings. During online finetuning, the base policy is frozen and RL learns only a residual perturbation on this latent, whose decoded representation conditions the frozen action generator. We instantiate ZPRL on flow-matching policies and evaluate it on eight simulation tasks and four real-world tasks. Across diverse manipulation settings, ZPRL improves both sample efficiency and final performance over strong post-training baselines. In the real world, ZPRL improves the average success rate on four tasks by 33.7% over imitation base policies while producing smoother exploration behaviors than an action residual counterpart. These results suggest that a compact, task-aligned bottleneck latent provides an effective interface for online RL adaptation. More videos can be found at https://manutdmoon.github.io/ZPRL/.

Shared Autonomy Assisted by Impedance-Driven Anisotropic Guidance Field

Shared autonomy (SA) enables robots to infer human intent and assist in its achievement. While most research focuses on improving intent inference, it overlooks whether humans can understand the robot's intent in return. Without such mutual understanding, collaboration becomes less effective, degrading user experience and task performance. To address this gap, previous studies have explicitly conveyed the robot intent through additional interfaces, which remain unintuitive and limited in expressiveness. Inspired by impedance control, we propose Impedance-Driven Anisotropic Guidance Field Enhanced Shared Autonomy (IAGF-SA), a novel paradigm that extends SA with an embodied, physically-grounded communication channel. This channel adaptively modulates the robot's dynamic response to human input, enabling intuitive, continuous, physically-grounded robot intent communication while naturally guiding human actions. User studies across three scenarios and two teleoperation interfaces indicate that IAGF-SA improves task performance, human-robot agreement, and subjective experience, thus demonstrating its effectiveness in enhancing human-robot communication and collaboration.

RichMap: A Reachability Map Balancing Precision, Efficiency, and Flexibility for Rich Robot Manipulation Tasks

This paper presents RichMap, a high-precision reachability map representation designed to balance efficiency and flexibility for versatile robot manipulation tasks. By refining the classic grid-based structure, we propose a streamlined approach that achieves performance close to compact map forms (e.g., RM4D) while maintaining structural flexibility. Our method utilizes theoretical capacity bounds on $\mathbb{S}^2$ (or $SO(3)$) to ensure rigorous coverage and employs an asynchronous pipeline for efficient construction. We validate the map against comprehensive metrics, pursuing high prediction accuracy ($>98\%$), low false positive rates ($1\sim2\%$), and fast large-batch query ($\sim$15 $μ$s/query). We extend the framework applications to quantify robot workspace similarity via maximum mean discrepancy (MMD) metrics and demonstrate energy-based guidance for diffusion policy transfer, achieving up to $26\%$ improvement for cross-embodiment scenarios in the block pushing experiment.

From Video to Control: A Survey of Learning Manipulation Interfaces from Temporal Visual Data

Video is a scalable observation of physical dynamics: it captures how objects move, how contact unfolds, and how scenes evolve under interaction -- all without requiring robot action labels. Yet translating this temporal structure into reliable robotic control remains an open challenge, because video lacks action supervision and differs from robot experience in embodiment, viewpoint, and physical constraints. This survey reviews methods that exploit non-action-annotated temporal video to learn control interfaces for robotic manipulation. We introduce an \emph{interface-centric taxonomy} organized by where the video-to-control interface is constructed and what control properties it enables, identifying three families: direct video--action policies, which keep the interface implicit; latent-action methods, which route temporal structure through a compact learned intermediate; and explicit visual interfaces, which predict interpretable targets for downstream control. For each family, we analyze control-integration properties -- how the loop is closed, what can be verified before execution, and where failures enter. A cross-family synthesis reveals that the most pressing open challenges center on the \emph{robotics integration layer} -- the mechanisms that connect video-derived predictions to dependable robot behavior -- and we outline research directions toward closing this gap.

The USTC-NERCSLIP Systems for the CHiME-9 MCoRec Challenge

This report details our submission to the CHiME-9 MCoRec Challenge on recognizing and clustering multiple concurrent natural conversations within indoor social settings. Unlike conventional meetings centered on a single shared topic, this scenario contains multiple parallel dialogues--up to eight speakers across up to four simultaneous conversations--with a speech overlap rate exceeding 90%. To tackle this, we propose a multimodal cascaded system that leverages per-speaker visual streams extracted from synchronized 360 degree video together with single-channel audio. Our system improves three components of the pipeline by leveraging enhanced audio-visual pretrained models: Active Speaker Detection (ASD), Audio-Visual Target Speech Extraction (AVTSE), and Audio-Visual Speech Recognition (AVSR). The AVSR module further incorporates Whisper and LLM techniques to boost transcription accuracy. Our best single cascaded system achieves a Speaker Word Error Rate (WER) of 32.44% on the development set. By further applying ROVER to fuse outputs from diverse front-end and back-end variants, we reduce Speaker WER to 31.40%. Notably, our LLM-based zero-shot conversational clustering achieves a speaker clustering F1 score of 1.0, yielding a final Joint ASR-Clustering Error Rate (JACER) of 15.70%.

Failure-Aware Bimanual Teleoperation via Conservative Value Guided Assistance

Teleoperation of high-precision manipulation is con-strained by tight success tolerances and complex contact dy-namics, which make impending failures difficult for human operators to anticipate under partial observability. This paper proposes a value-guided, failure-aware framework for bimanual teleoperation that provides compliant haptic assistance while pre-serving continuous human authority. The framework is trained entirely from heterogeneous offline teleoperation data containing both successful and failed executions. Task feasibility is mod-eled as a conservative success score learned via Conservative Value Learning, yielding a risk-sensitive estimate that remains reliable under distribution shift. During online operation, the learned success score regulates the level of assistance, while a learned actor provides a corrective motion direction. Both are integrated through a joint-space impedance interface on the master side, yielding continuous guidance that steers the operator away from failure-prone actions without overriding intent. Experimental results on contact-rich manipulation tasks demonstrate improved task success rates and reduced operator workload compared to conventional teleoperation and shared-autonomy baselines, indicating that conservative value learning provides an effective mechanism for embedding failure awareness into bilateral teleoperation. Experimental videos are available at https://www.youtube.com/watch?v=XDTsvzEkDRE

UniBiDex: A Unified Teleoperation Framework for Robotic Bimanual Dexterous Manipulation

We present UniBiDex a unified teleoperation framework for robotic bimanual dexterous manipulation that supports both VRbased and leaderfollower input modalities UniBiDex enables realtime contactrich dualarm teleoperation by integrating heterogeneous input devices into a shared control stack with consistent kinematic treatment and safety guarantees The framework employs nullspace control to optimize bimanual configurations ensuring smooth collisionfree and singularityaware motion across tasks We validate UniBiDex on a longhorizon kitchentidying task involving five sequential manipulation subtasks demonstrating higher task success rates smoother trajectories and improved robustness compared to strong baselines By releasing all hardware and software components as opensource we aim to lower the barrier to collecting largescale highquality human demonstration datasets and accelerate progress in robot learning.

iFlyBot-VLM Technical Report

We introduce iFlyBot-VLM, a general-purpose Vision-Language Model (VLM) used to improve the domain of Embodied Intelligence. The central objective of iFlyBot-VLM is to bridge the cross-modal semantic gap between high-dimensional environmental perception and low-level robotic motion control. To this end, the model abstracts complex visual and spatial information into a body-agnostic and transferable Operational Language, thereby enabling seamless perception-action closed-loop coordination across diverse robotic platforms. The architecture of iFlyBot-VLM is systematically designed to realize four key functional capabilities essential for embodied intelligence: 1) Spatial Understanding and Metric Reasoning; 2) Interactive Target Grounding; 3) Action Abstraction and Control Parameter Generation; 4) Task Planning and Skill Sequencing. We envision iFlyBot-VLM as a scalable and generalizable foundation model for embodied AI, facilitating the progression from specialized task-oriented systems toward generalist, cognitively capable agents. We conducted evaluations on 10 current mainstream embodied intelligence-related VLM benchmark datasets, such as Blink and Where2Place, and achieved optimal performance while preserving the model's general capabilities. We will publicly release both the training data and model weights to foster further research and development in the field of Embodied Intelligence.