arXiv subjects

Zixing Wang

Publications and source records attributed to Zixing Wang.

At least 19 recordsLinked to original sources

arXiv2026

SkillFab: An Agent-Native Skill Production Platform

SkillFab is an agent-native platform for turning missing capabilities into reviewed, reusable Agent Skills. At runtime, agents first search for reusable skills; when no adequate skill exists, the unmet capability becomes a demand-first issue before any repository or implementation branch needs to exist. Development then proceeds through a SkillFab-managed repository, Git-ingested commit evidence, maintainer review, and registry publication. The same lifecycle is exposed through web, REST, and MCP surfaces, so humans, scripts, and external agents operate on shared state rather than separate task logs. The current system uses scoped Git push URLs, native range commit ingestion, workflow-state reads, and workflow-event histories to make long-running agent work reviewable and recoverable. We document the platform model, architecture, implemented capabilities, and three case studies: an end-to-end OS-detect skill run, a Docker research package that converts operational practice into reusable skill knowledge, and an external optimization case showing how improved skill artifacts can enter SkillFab as reviewable, versioned submissions. Deployment: https://skillfab.ai.

cs.SE↗

arXiv2026

Efficient Sim-to-Real Transfer of World-Action Models from Synthetic Priors

Bridging the sim-to-real gap is a core challenge in deploying learned manipulation policies. Sim-to-real learning is attractive because it can replace expensive real robot demonstrations with scalable synthetic data, yet world-action models have not previously been shown to transfer from simulation to real robotic manipulation. We study whether a world-action model can be trained from synthetic priors and deployed zero-shot in the real world. To this end, we build upon Cosmos Policy, a video diffusion model adapted for visuomotor control. We construct simulation environments with extensive domain randomization and generate demonstrations using the AnyTask motion planning pipeline. We evaluate our approach across object lifting, drawer opening, and pick-and-place tasks using ${\sim}800$ synthetic demonstrations per task and no real demonstrations. When deployed zero-shot on a Franka Robot, our policy attains a 35\% average success rate. To our knowledge, this represents the first successful sim-to-real transfer of a world-action model for robotic manipulation.

cs.RO↗

arXiv2026

Passive iFIR filters for data-driven velocity control in robotics

We present a passive, data-driven velocity control method for nonlinear robotic manipulators that achieves better tracking performance than optimized PID with comparable design complexity. Using only three minutes of probing data, a VRFT-based design identifies passive iFIR controllers that (i) preserve closed-loop stability via passivity constraints and (ii) outperform a VRFT-tuned PID baseline on the Franka Research 3 robot in both joint-space and Cartesian-space velocity control, achieving up to a 74.5% reduction in tracking error for the Cartesian velocity tracking experiment with the most demanding reference model. When the robot end-effector dynamics change, the controller can be re-learned from new data, regaining nominal performance. This study bridges learning-based control and stability-guaranteed design: passive iFIR learns from data while retaining passivity-based stability guarantees, unlike many learning-based approaches.

cs.RO↗

arXiv2026

SkillTester: Benchmarking Utility and Security of Agent Skills

This technical report presents SkillTester, a tool for evaluating the utility and security of agent skills. Its evaluation framework combines paired baseline and with-skill execution conditions with a separate security probe suite. Grounded in a comparative utility principle and a user-facing simplicity principle, the framework normalizes raw execution artifacts into a utility score, a security score, and a three-level security status label. More broadly, it can be understood as a comparative quality-assurance harness for agent skills in an agent-first world. The public service is deployed at https://skilltester.ai, and the broader project is maintained at https://github.com/skilltester-ai/skilltester.

cs.CR↗

arXiv2026

ExpertGen: Scalable Sim-to-Real Expert Policy Learning from Imperfect Behavior Priors

Learning generalizable and robust behavior cloning policies requires large volumes of high-quality robotics data. While human demonstrations (e.g., through teleoperation) serve as the standard source for expert behaviors, acquiring such data at scale in the real world is prohibitively expensive. This paper introduces ExpertGen, a framework that automates expert policy learning in simulation to enable scalable sim-to-real transfer. ExpertGen first initializes a behavior prior using a diffusion policy trained on imperfect demonstrations, which may be synthesized by large language models or provided by humans. Reinforcement learning is then used to steer this prior toward high task success by optimizing the diffusion model's initial noise while keep original policy frozen. By keeping the pretrained diffusion policy frozen, ExpertGen regularizes exploration to remain within safe, human-like behavior manifolds, while also enabling effective learning with only sparse rewards. Empirical evaluations on challenging manipulation benchmarks demonstrate that ExpertGen reliably produces high-quality expert policies with no reward engineering. On industrial assembly tasks, ExpertGen achieves a 90.5% overall success rate, while on long-horizon manipulation tasks it attains 85% overall success, outperforming all baseline methods. The resulting policies exhibit dexterous control and remain robust across diverse initial configurations and failure states. To validate sim-to-real transfer, the learned state-based expert policies are further distilled into visuomotor policies via DAgger and successfully deployed on real robotic hardware.

cs.RO↗

arXiv2026

PPGuide: Steering Diffusion Policies with Performance Predictive Guidance

Diffusion policies have shown to be very efficient at learning complex, multi-modal behaviors for robotic manipulation. However, errors in generated action sequences can compound over time which can potentially lead to failure. Some approaches mitigate this by augmenting datasets with expert demonstrations or learning predictive world models which might be computationally expensive. We introduce Performance Predictive Guidance (PPGuide), a lightweight, classifier-based framework that steers a pre-trained diffusion policy away from failure modes at inference time. PPGuide makes use of a novel self-supervised process: it uses attention-based multiple instance learning to automatically estimate which observation-action chunks from the policy's rollouts are relevant to success or failure. We then train a performance predictor on this self-labeled data. During inference, this predictor provides a real-time gradient to guide the policy toward more robust actions. We validated our proposed PPGuide across a diverse set of tasks from the Robomimic and MimicGen benchmarks, demonstrating consistent improvements in performance.

cs.RO↗

arXiv2025

FlowDet: Overcoming Perspective and Scale Challenges in Real-Time End-to-End Traffic Detection

End-to-end object detectors offer a promising NMS-free paradigm for real-time applications, yet their high computational cost remains a significant barrier, particularly for complex scenarios like intersection traffic monitoring. To address this challenge, we propose FlowDet, a high-speed detector featuring a decoupled encoder optimization strategy applied to the DETR architecture. Specifically, FlowDet employs a novel Geometric Deformable Unit (GDU) for traffic-aware geometric modeling and a Scale-Aware Attention (SAA) module to maintain high representational power across extreme scale variations. To rigorously evaluate the model's performance in environments with severe occlusion and high object density, we collected the Intersection-Flow-5k dataset, a new challenging scene for this task. Evaluated on Intersection-Flow-5k, FlowDet establishes a new state-of-the-art. Compared to the strong RT-DETR baseline, it improves AP(test) by 1.5% and AP50(test) by 1.6%, while simultaneously reducing GFLOPs by 63.2% and increasing inference speed by 16.2%. Our work demonstrates a new path towards building highly efficient and accurate detectors for demanding, real-world perception systems. The Intersection-Flow-5k dataset is available at https://github.com/AstronZh/Intersection-Flow-5K.

cs.CV↗

arXiv2025

Passive Lifted FIR Filters for Nonlinear System Identification

Passivity is a fundamental property of physical systems. In data-driven modeling, ensuring that a learned model preserves this structural property is critical to avoiding instability in close loop. Although linear passive system identification is well-established, nonlinear extensions remain challenging. We propose nonlinear operators defined through passivity-preserving lifting of linear passive FIR filters. Passivity is enforced efficiently through frequency-domain constraints, and the nonlinear lifting includes output feedback for expressivity. Numerical and real-world experiments demonstrate the framework capabilities, including the computational advantage of frequency-domain constraints against LMI-based alternatives.

eess.SY↗

arXiv2025

Multimodal Human-Intent Modeling for Contextual Robot-to-Human Handovers of Arbitrary Objects

Human-robot object handover is a crucial element for assistive robots that aim to help people in their daily lives, including elderly care, hospitals, and factory floors. The existing approaches to solving these tasks rely on pre-selected target objects and do not contextualize human implicit and explicit preferences for handover, limiting natural and smooth interaction between humans and robots. These preferences can be related to the target object selection from the cluttered environment and to the way the robot should grasp the selected object to facilitate desirable human grasping during handovers. Therefore, this paper presents a unified approach that selects target distant objects using human verbal and non-verbal commands and performs the handover operation by contextualizing human implicit and explicit preferences to generate robot grasps and compliant handover motion sequences. We evaluate our integrated framework and its components through real-world experiments and user studies with arbitrary daily-life objects. The results of these evaluations demonstrate the effectiveness of our proposed pipeline in handling object handover tasks by understanding human preferences. Our demonstration videos can be found at https://youtu.be/6z27B2INl-s.

cs.RO↗

arXiv2025

Physics-Conditioned Grasping for Stable Tool Use

Tool use often fails not because robots misidentify tools, but because grasps cannot withstand task-induced wrench. Existing vision-language manipulation systems ground tools and contact regions from language yet select grasps under quasi-static or geometry-only assumptions. During interaction, inertial impulse and lever-arm amplification generate wrist torque and tangential loads that trigger slip and rotation. We introduce inverse Tool-use Planning (iTuP), which selects grasps by minimizing predicted interaction wrench along a task-conditioned trajectory. From rigid-body mechanics, we derive torque, slip, and alignment penalties, and train a Stable Dynamic Grasp Network (SDG-Net) to approximate these trajectory-conditioned costs for real-time scoring. Across hammering, sweeping, knocking, and reaching in simulation and on hardware, SDG-Net suppresses induced torque up to 17.6%, shifts grasps below empirically observed instability thresholds, and improves real-world success by 17.5% over a compositional baseline. Improvements concentrate where wrench amplification dominates, showing that robot tool use requires wrench-aware grasp selection, not perception alone.

cs.RO↗

arXiv2025

Implicit Physics-aware Policy for Dynamic Manipulation of Rigid Objects via Soft Body Tools

Recent advancements in robot tool use have unlocked their usage for novel tasks, yet the predominant focus is on rigid-body tools, while the investigation of soft-body tools and their dynamic interaction with rigid bodies remains unexplored. This paper takes a pioneering step towards dynamic one-shot soft tool use for manipulating rigid objects, a challenging problem posed by complex interactions and unobservable physical properties. To address these problems, we propose the Implicit Physics-aware (IPA) policy, designed to facilitate effective soft tool use across various environmental configurations. The IPA policy conducts system identification to implicitly identify physics information and predict goal-conditioned, one-shot actions accordingly. We validate our approach through a challenging task, i.e., transporting rigid objects using soft tools such as ropes to distant target positions in a single attempt under unknown environment physics parameters. Our experimental results indicate the effectiveness of our method in efficiently identifying physical properties, accurately predicting actions, and smoothly generalizing to real-world environments. The related video is available at: https://youtu.be/4hPrUDTc4Rg?si=WUZrT2vjLMt8qRWA

cs.RO↗

arXiv2024

Dissipative iFIR filters for data-driven design

We tackle the problem of providing closed-loop stability guarantees with a scalable data-driven design. We combine virtual reference feedback tuning with dissipativity constraints on the controller for closed-loop stability. The constraints are formulated as a set of linear inequalities in the frequency domain. This leads to a convex problem that is scalable with respect to the length of the data and the complexity of the controller. An extension of virtual reference feedback tuning to include disturbance dynamics is also discussed. The proposed data-driven control design is illustrated by a soft gripper impedance control example.

eess.SY↗

arXiv2024

Passive iFIR Filters for Data-Driven Control

We consider the design of a new class of passive iFIR controllers given by the parallel action of an integrator and a finite impulse response filter. iFIRs are more expressive than PID controllers but retain their features and simplicity. The paper provides a model-free data-driven design for passive iFIR controllers based on virtual reference feedback tuning. Passivity is enforced through constrained optimization (three different formulations are discussed). The proposed design does not rely on large datasets or accurate plant models.

eess.SY↗

arXiv2023

AnyPose: Anytime 3D Human Pose Forecasting via Neural Ordinary Differential Equations

Anytime 3D human pose forecasting is crucial to synchronous real-world human-machine interaction, where the term ``anytime" corresponds to predicting human pose at any real-valued time step. However, to the best of our knowledge, all the existing methods in human pose forecasting perform predictions at preset, discrete time intervals. Therefore, we introduce AnyPose, a lightweight continuous-time neural architecture that models human behavior dynamics with neural ordinary differential equations. We validate our framework on the Human3.6M, AMASS, and 3DPW dataset and conduct a series of comprehensive analyses towards comparison with existing methods and the intersection of human pose and neural ordinary differential equations. Our results demonstrate that AnyPose exhibits high-performance accuracy in predicting future poses and takes significantly lower computational time than traditional methods in solving anytime prediction tasks.

cs.CV↗

arXiv2023

DeRi-IGP: Learning to Manipulate Rigid Objects Using Deformable Objects via Iterative Grasp-Pull

Robotic manipulation of rigid objects via deformable linear objects (DLO) such as ropes is an emerging field of research with applications in various rigid object transportation tasks. A few methods that exist in this field suffer from limited robot action and operational space, poor generalization ability, and expensive model-based development. To address these challenges, we propose a universally applicable moving primitive called Iterative Grasp-Pull (IGP). We also introduce a novel vision-based neural policy that learns to parameterize the IGP primitive to manipulate DLO and transport their attached rigid objects to the desired goal locations. Additionally, our decentralized algorithm design allows collaboration among multiple agents to manipulate rigid objects using DLO. We evaluated the effectiveness of our approach in both simulated and real-world environments for a variety of soft-rigid body manipulation tasks. In the real world, we also demonstrate the effectiveness of our decentralized approach through human-robot collaborative transportation of rigid objects to given goal locations. We also showcase the large operational space of IGP primitive by solving distant object acquisition tasks. Lastly, we compared our approach with several model-based and learning-based baseline methods. The results indicate that our method surpasses other approaches by a significant margin.

cs.RO↗

arXiv2023

Efficient Q-Learning over Visit Frequency Maps for Multi-agent Exploration of Unknown Environments

The robot exploration task has been widely studied with applications spanning from novel environment mapping to item delivery. For some time-critical tasks, such as rescue catastrophes, the agent is required to explore as efficiently as possible. Recently, Visit Frequency-based map representation achieved great success in such scenarios by discouraging repetitive visits with a frequency-based penalty. However, its relatively large size and single-agent settings hinder its further development. In this context, we propose Integrated Visit Frequency Map, which encodes identical information as Visit Frequency Map with a more compact size, and a visit frequency-based multi-agent information exchange and control scheme that is able to accommodate both representations. Through tests in diverse settings, the results indicate our proposed methods can achieve a comparable level of performance of VFM with lower bandwidth requirements and generalize well to different multi-agent setups including real-world environments.

cs.RO↗

arXiv2023

DeRi-Bot: Learning to Collaboratively Manipulate Rigid Objects via Deformable Objects

Recent research efforts have yielded significant advancements in manipulating objects under homogeneous settings where the robot is required to either manipulate rigid or deformable (soft) objects. However, the manipulation under heterogeneous setups that involve both rigid and one-dimensional (1D) deformable objects remains an unexplored area of research. Such setups are common in various scenarios that involve the transportation of heavy objects via ropes, e.g., on factory floors, at disaster sites, and in forestry. To address this challenge, we introduce DeRi-Bot, the first framework that enables the collaborative manipulation of rigid objects with deformable objects. Our framework comprises an Action Prediction Network (APN) and a Configuration Prediction Network (CPN) to model the complex pattern and stochasticity of soft-rigid body systems. We demonstrate the effectiveness of DeRi-Bot in moving rigid objects to a target position with ropes connected to robotic arms. Furthermore, DeRi-Bot is a distributive method that can accommodate an arbitrary number of robots or human partners without reconfiguration or retraining. We evaluate our framework in both simulated and real-world environments and show that it achieves promising results with strong generalization across different types of objects and multi-agent settings, including human-robot collaboration.

cs.RO↗

arXiv2021

A Generalized Determinant of Matrices and Applications

A generalized definition of the determinant of matrices is given, which is compatible with the usual determinant for square matrices and keeps many important properties, such as being an alternating multilinear function, keeping multiplication formula and partly keeping the Cauchy-Binet's formula. As applications of the new theory, the generalized Cramer's rule and the generalized oriented volume are obtained.

math.CA↗