arXiv subjects

Junheng Li

Publications and source records attributed to Junheng Li.

At least 19 recordsLinked to original sources

arXiv2026

Shooting for Contact: Contact-Implicit Multiple Shooting for Dynamic Motion Retargeting

Motion retargeting approaches often prioritize kinematic similarity over whole-body dynamics, contact consistency, and actuation limits, yielding references that are difficult for reinforcement learning (RL) policies to reproduce, particularly for contact-rich behaviors. We present a contact-implicit, direct simulation-based multiple shooting (DSMS) framework that transforms kinematically feasible references into dynamically feasible whole-body trajectories. By embedding a differentiable simulator within a nonlinear program, DSMS resolves contact, friction, impacts, self-collision, and joint limits internally while enforcing tracking, actuation, and task constraints without prescribing a contact schedule or introducing explicit contact constraints. Compared with existing retargeting methods, DSMS accelerates motion-imitation RL training and yields policies with high success rates and low tracking error. We further demonstrate zero-shot sim-to-real transfer on the Unitree G1 through command-conditioned contact-rich crawling and a highly dynamic 180-degree jump-turn.

cs.RO↗

arXiv2026

PAC-MAN: Perception-Aware CBF-RL for Whole-Body Safety in Humanoid Dodgeball

We present PAC-MAN, a perception-aware CBF-RL framework that couples control-barrier safety with deployment-realistic onboard sensing for whole-body humanoid dodgeball. The deployed policy sees the ball only as segmentation-masked depth from a head-mounted camera, while training-time CBF guidance represents clearance to every body link, and an adversarial motion prior regularizes the resulting evasive reflexes. We evaluate on a controlled any-link contact benchmark with seeded throws in two regimes: single throws and a deployment loop in which the robot walks back to its station and recovers between throws. On this benchmark, the policy comes within a few points of a privileged state oracle: a fixed onboard camera alone is adequate for evasion. We find that usable barrier structure depends on perceptual observability: Joint-CBF gives the best performance with accurate ball states, degrades under fixed-camera observations when used only as training guidance, and recovers with a ball-tracking gimbal or privileged runtime filter. We therefore deploy a lightweight Link-CBF policy zero-shot on the Unitree G1 in the real world, where it tolerates imperfect perception, succeeds on 95% of throws, and uses semantic segmentation to dodge different balls.

cs.RO↗

arXiv2026

Accelerating and Scaling MPC-Guided Reinforcement Learning for Humanoid Locomotion and Manipulation

In humanoid motion control, model predictive control (MPC) offers physically grounded prediction and constraint handling, while reinforcement learning (RL) enables robust whole-body skills through large-scale simulation. However, using MPC inside RL often requires time-consuming problem construction or excessive training overhead, making such frameworks difficult to justify in practice. This work studies efficient training-time MPC guidance for humanoid locomotion and manipulation, termed MPC-RL. We introduce a centroidal-dynamics MPC reward formulation that leverages guidance from MPC trajectories in training time. To make this practical in massively parallel RL, we develop $\pi^n$MPC, a parallel-in-horizon and construction-free batched GPU MPC solver that operates directly on time-varying dynamics to avoid high memory usage and pre-compilation. Through a variety of comparative studies and hardware validations, we have found that MPC-RL achieves superior performance in locomotion and manipulation skills. The code base is available at https://github.com/junhengl/mpc-rl.

cs.RO↗

arXiv2026

HANDOFF: Humanoid Agentic Task-Space Whole-Body Control via Distilled Complementary Teachers

For a humanoid robot to be deployed in the real world, the choice of command space (i.e., the interface between task planning and whole-body control) is crucial. Existing whole-body controllers typically demand dense kinematic or spatial references that planners struggle to synthesize from task semantics. We instead propose a compact, explicit interface that is intuitive, general, modular, and expressive enough for diverse loco-manipulation skills. To this end, we introduce HANDOFF, a single humanoid whole-body controller that follows this interface and is distilled via multi-teacher KL distillation under a context-conditioned gating scheme into a mixture-of-experts student from three complementary specialists: whole-body motion tracking with safety-filtered data, locomotion, and fall-recovery. On the Unitree G1, HANDOFF matches state-of-the-art velocity tracking and offers one of the largest robust manipulation workspaces. We further demonstrate hardware feasibility through multiple natural-language-driven task roll-outs, powered by a VLM-driven agentic planner with no task-specific data or controller fine-tuning.

cs.RO↗

arXiv2026

See Before You Code: Learning Visual Priors for Spatially Aware Educational Animation Generation

Large language models can generate executable code for educational animations, but the resulting renders often exhibit visual defects, including element overlap, misalignment, and broken animation continuity. These defects cannot be reliably detected from the code alone and become apparent only after execution. We formalize this problem as render-feedback-aware constrained code generation: given a natural language specification, the model must generate executable code whose rendered output satisfies structured quality criteria that can be evaluated only after rendering. To address this problem, we introduce OmniManim, a render-feedback-aware educational animation generation framework built around a shared scene state, explicit visual planning, structured post-render diagnostics, and localized repair. Within OmniManim, the Vision Agent is a task-specific visual planning module: it predicts sparse keyframe layouts with coarse-to-fine bounding-box denoising and optimizes an interpolation-aware objective to reduce intermediate-frame failures induced by downstream animation interpolation. We further construct two datasets, ManimLayout-1K and EduRequire-500, and provide a reproducible evaluation protocol covering executability, instructional quality, visual quality, and efficiency. On EduRequire-500, OmniManim improves measured render quality over both single-model baselines and existing multi-agent frameworks. Systematic ablation studies further verify that explicit visual planning, especially its coarse spatial prior, bounding-box refinement, and interpolation-aware optimization, is central to these gains.

cs.AI↗

arXiv2026

On Surprising Effects of Risk-Aware Domain Randomization for Contact-Rich Sampling-based Predictive Control

Domain randomization (DR) is widely used in policy learning to improve robustness to modeling error, but remains underexplored in contact-rich sampling-based predictive control (SPC), where rollout quality is highly sensitive to uncertainty. In this work, we take the first step by studying risk-aware DR in predictive sampling on a simple yet representative Push-T task, comparing average, optimistic, and pessimistic rollout aggregations under randomized model instances. Our initial results suggest that DR affects not only robustness to model error, but also the effective cost landscape seen by the sampling-based optimizer, by reshaping the basin of attraction around contact-producing actions. This opens up potential for exploring better grounded risk-aware contact-rich SPC under model uncertainty. Video: https://youtu.be/f1F0ALXxhSM

cs.RO↗

arXiv2026

MIRROR: Visual Motion Imitation via Real-time Retargeting and Teleoperation with Parallel Differential Inverse Kinematics

Real-time humanoid teleoperation requires inverse kinematics (IK) solvers that are both responsive and constraint-safe under kinematic redundancy and self-collision constraints. While differential IK enables efficient online retargeting, its locally linearized updates are inherently basin-dependent and often become trapped near joint limits, singularities, or active collision boundaries, leading to unsafe or stagnant behavior. We propose a GPU-parallelized, continuation-based differential IK that improves escape from such constraint-induced local minima while preserving real-time performance, promoting safety and stability. Multiple constrained IK quadratic programs are evaluated in parallel, together with a self-collision avoidance control barrier function (CBF), and a Lyapunov-based progression criterion selects updates that reduce the final global task-space error. The method is paired with a visual skeletal pose estimation pipeline that enables robust, real-time upper-body teleoperation on the THEMIS humanoid robot hardware in real-world tasks.

cs.RO↗

arXiv2026

$\pi$MPC: A Parallel-in-horizon and Construction-free NMPC Solver

The alternating direction method of multipliers (ADMM) has gained increasing popularity in embedded model predictive control (MPC) due to its code simplicity and pain-free parameter selection. However, existing ADMM solvers either target general quadratic programming (QP) problems or exploit sparse MPC formulations via Riccati recursions, which are inherently sequential and therefore difficult to parallelize for long prediction horizons. This technical note proposes a novel \textit{parallel-in-horizon} and \textit{construction-free} nonlinear MPC algorithm, termed $\pi$MPC, which combines a new variable-splitting scheme with a velocity-based system representation in the ADMM framework, enabling horizon-wise parallel execution while operating directly on system matrices without explicit MPC-to-QP construction. Numerical experiments and accompanying code are provided to validate the effectiveness of the proposed method.

math.OC↗

arXiv2026

Walk the PLANC: Physics-Guided RL for Agile Humanoid Locomotion on Constrained Footholds

Bipedal humanoid robots must precisely coordinate balance, timing, and contact decisions when locomoting on constrained footholds such as stepping stones, beams, and planks -- even minor errors can lead to catastrophic failure. Classical optimization and control pipelines handle these constraints well but depend on highly accurate mathematical representations of terrain geometry, making them prone to error when perception is noisy or incomplete. Meanwhile, reinforcement learning has shown strong resilience to disturbances and modeling errors, yet end-to-end policies rarely discover the precise foothold placement and step sequencing required for discontinuous terrain. These contrasting limitations motivate approaches that guide learning with physics-based structure rather than relying purely on reward shaping. In this work, we introduce a locomotion framework in which a reduced-order stepping planner supplies dynamically consistent motion targets that steer the RL training process via Control Lyapunov Function (CLF) rewards. This combination of structured footstep planning and data-driven adaptation produces accurate, agile, and hardware-validated stepping-stone locomotion on a humanoid robot, substantially improving reliability compared to conventional model-free reinforcement-learning baselines.

cs.RO↗

arXiv2025

DiffCoTune: Differentiable Co-Tuning for Cross-domain Robot Control

The deployment of robot controllers is hindered by modeling discrepancies due to necessary simplifications for computational tractability or inaccuracies in data-generating simulators. Such discrepancies typically require ad-hoc tuning to meet the desired performance, thereby ensuring successful transfer to a target domain. We propose a framework for automated, gradient-based tuning to enhance performance in the deployment domain by leveraging differentiable simulators. Our method collects rollouts in an iterative manner to co-tune the simulator and controller parameters, enabling systematic transfer within a few trials in the deployment domain. Specifically, we formulate multi-step objectives for tuning and employ alternating optimization to effectively adapt the controller to the deployment domain. The scalability of our framework is demonstrated by co-tuning model-based and learning-based controllers of arbitrary complexity for tasks ranging from low-dimensional cart-pole stabilization to high-dimensional quadruped and biped tracking, showing performance improvements across different deployment domains.

cs.RO↗

arXiv2025

Gait-Net-augmented Implicit Kino-dynamic MPC for Dynamic Variable-frequency Humanoid Locomotion over Discrete Terrains

Reduced-order-model-based optimal control techniques for humanoid locomotion struggle to adapt step duration and placement simultaneously in dynamic walking gaits due to their reliance on fixed-time discretization, which limits responsiveness to various disturbances and results in suboptimal performance in challenging conditions. In this work, we propose a Gait-Net-augmented implicit kino-dynamic model-predictive control (MPC) to simultaneously optimize step location, step duration, and contact forces for natural variable-frequency locomotion. The proposed method incorporates a Gait-Net-augmented Sequential Convex MPC algorithm to solve multi-linearly constrained variables by iterative quadratic programs. At its core, a lightweight Gait-frequency Network (Gait-Net) determines the preferred step duration in terms of variable MPC sampling times, simplifying step duration optimization to the parameter level. Additionally, it enhances and updates the spatial reference trajectory within each sequential iteration by incorporating local solutions, allowing the projection of kinematic constraints to the design of reference trajectories. We validate the proposed algorithm in high-fidelity simulations and on small-size humanoid hardware, demonstrating its capability for variable-frequency and 3-D discrete terrain locomotion with only a one-step preview of terrain data.

cs.RO↗

arXiv2025

Humanoid Locomotion and Manipulation: Current Progress and Challenges in Control, Planning, and Learning

Humanoid robots hold great potential to perform various human-level skills, involving unified locomotion and manipulation in real-world settings. Driven by advances in machine learning and the strength of existing model-based approaches, these capabilities have progressed rapidly, but often separately. This survey offers a comprehensive overview of the state-of-the-art in humanoid locomotion and manipulation (HLM), with a focus on control, planning, and learning methods. We first review the model-based methods that have been the backbone of humanoid robotics for the past three decades. We discuss contact planning, motion planning, and whole-body control, highlighting the trade-offs between model fidelity and computational efficiency. Then the focus is shifted to examine emerging learning-based methods, with an emphasis on reinforcement and imitation learning that enhance the robustness and versatility of loco-manipulation skills. Furthermore, we assess the potential of integrating foundation models with humanoid embodiments to enable the development of generalist humanoid agents. This survey also highlights the emerging role of tactile sensing, particularly whole-body tactile feedback, as a crucial modality for handling contact-rich interactions. Finally, we compare the strengths and limitations of model-based and learning-based paradigms from multiple perspectives, such as robustness, computational efficiency, versatility, and generalizability, and suggest potential solutions to existing challenges.

cs.RO↗

arXiv2024

Autotuning Bipedal Locomotion MPC with GRFM-Net for Efficient Sim-to-Real Transfer

Bipedal locomotion control is essential for humanoid robots to navigate complex, human-centric environments. While optimization-based control designs are popular for integrating sophisticated models of humanoid robots, they often require labor-intensive manual tuning. In this work, we address the challenges of parameter selection in bipedal locomotion control using DiffTune, a model-based autotuning method that leverages differential programming for efficient parameter learning. A major difficulty lies in balancing model fidelity with differentiability. We address this difficulty using a low-fidelity model for differentiability, enhanced by a Ground Reaction Force-and-Moment Network (GRFM-Net) to capture discrepancies between MPC commands and actual control effects. We validate the parameters learned by DiffTune with GRFM-Net in hardware experiments, which demonstrates the parameters' optimality in a multi-objective setting compared with baseline parameters, reducing the total loss by up to 40.5$\%$ compared with the expert-tuned parameters. The results confirm the GRFM-Net's effectiveness in mitigating the sim-to-real gap, improving the transferability of simulation-learned parameters to real hardware.

cs.RO↗

arXiv2024

Continuous Dynamic Bipedal Jumping via Real-time Variable-model Optimization

Dynamic and continuous jumping remains an open yet challenging problem in bipedal robot control. Real-time planning with full body dynamics over the entire jumping trajectory presents unsolved challenges in computation burden. In this paper, we propose a novel variable-model optimization approach, a unified framework of variable-model trajectory optimization (TO) and variable-frequency Model Predictive Control (MPC), to effectively realize continuous and robust jumping planning and control on HECTOR bipedal robot in real-time. The proposed TO fuses variable-fidelity dynamics modeling of bipedal jumping motion in different jumping phases to balance trajectory accuracy and real-time computation efficiency. In addition, conventional fixed-frequency control approaches suffer from unsynchronized sampling frequencies, leading to mismatched modeling resolutions. We address this by aligning the MPC sampling frequency with the variable-model TO trajectory resolutions across different phases. In hardware experiments, we have demonstrated robust and dynamic jumps covering a distance of up to 40 cm (57% of robot height). To verify the repeatability of this experiment, we run 53 jumping experiments and achieve 90% success rate. In continuous jumps, we demonstrate continuous bipedal jumping with terrain height perturbations (up to 5 cm) and discontinuities (up to 20 cm gap).

cs.RO↗

arXiv2024

Adapting Gait Frequency for Posture-regulating Humanoid Push-recovery via Hierarchical Model Predictive Control

Current humanoid push-recovery strategies often use whole-body motion, yet they tend to overlook posture regulation. For instance, in manipulation tasks, the upper body may need to stay upright and have minimal recovery displacement. This paper introduces a novel approach to enhancing humanoid push-recovery performance under unknown disturbances and regulating body posture by tailoring the recovery stepping strategy. We propose a hierarchical-MPC-based scheme that analyzes and detects instability in the prediction window and quickly recovers through adapting gait frequency. Our approach integrates a high-level nonlinear MPC, a posture-aware gait frequency adaptation planner, and a low-level convex locomotion MPC. The planners predict the center of mass (CoM) state trajectories that can be assessed for precursors of potential instability and posture deviation. In simulation, we demonstrate improved maximum recoverable impulse by 131% on average compared with baseline approaches. In hardware experiments, a 125 ms advancement in recovery stepping timing/reflex has been observed with the proposed approach. We also demonstrate improved push-recovery performance and minimized body attitude change under 0.2 rad.

cs.RO↗

arXiv2023

Dynamic Loco-manipulation on HECTOR: Humanoid for Enhanced ConTrol and Open-source Research

Despite their remarkable advancement in locomotion and manipulation, humanoid robots remain challenged by a lack of synchronized loco-manipulation control, hindering their full dynamic potential. In this work, we introduce a versatile and effective approach to controlling and generalizing dynamic locomotion and loco-manipulation on humanoid robots via a Force-and-moment-based Model Predictive Control (MPC). Specifically, we proposed a simplified rigid body dynamics (SRBD) model to take into account both humanoid and object dynamics for humanoid loco-manipulation. This linear dynamics model allows us to directly solve for ground reaction forces and moments via an MPC problem to achieve highly dynamic real-time control. Our proposed framework is highly versatile and generalizable. We introduce HECTOR (Humanoid for Enhanced ConTrol and Open-source Research) platform to demonstrate its effectiveness in hardware experiments. With the proposed framework, HECTOR can maintain exceptional balance during double-leg stance mode, even when subjected to external force disturbances to the body or foot location. In addition, it can execute 3-D dynamic walking on a variety of uneven terrains, including wet grassy surfaces, slopes, randomly placed wood slats, and stacked wood slats up to 6 cm high with the speed of 0.6 m/s. In addition, we have demonstrated dynamic humanoid loco-manipulation over uneven terrain, carrying 2.5 kg load. HECTOR simulations, along with the proposed control framework, are made available as an open-source project. (https://github.com/DRCL-USC/Hector_Simulation).

cs.RO↗

arXiv2023

Kinodynamics-based Pose Optimization for Humanoid Loco-manipulation

This paper presents a novel approach for controlling humanoid robots to push heavy objects. The approach combines kinodynamics-based pose optimization and loco-manipulation model predictive control (MPC). The proposed pose optimization considers the object-robot dynamics model, robot kinematic constraints, and object parameters to plan the optimal pushing pose for the robot. The loco-manipulation MPC is used to track the optimal pose by coordinating pushing and ground reaction forces, ensuring accurate manipulation and stable locomotion. Numerical validation demonstrates the effectiveness of the framework, enabling the humanoid robot to push objects with various parameter setups. The pose optimization can be solved as a nonlinear programming (NLP) problem within an average of 250 ms. The proposed control scheme allows the humanoid robot to push objects weighing up to 20 kg (118$\%$ of the robot's mass). Additionally, it can recover the system from a 120 N lateral force disturbance applied for 0.3 s.

cs.RO↗

arXiv2023

Multi-contact MPC for Dynamic Loco-manipulation on Humanoid Robots

This paper presents a novel method to control humanoid robot dynamic loco-manipulation with multiple contact modes via multi-contact Model Predictive Control (MPC) framework. The proposed framework includes a multi-contact dynamics model capable of capturing various contact modes in loco-manipulation, such as hand-object contact and foot-ground contacts. Our proposed dynamics model represents the object dynamics as an external force acting on the system, which simplifies the model and makes it feasible for solving the MPC problem. In numerical validations, our multi-contact MPC framework only needs contact timings of each task and desired states to give MPC the knowledge of changes in contact modes in the prediction horizons in loco-manipulation. The proposed framework can control the humanoid robot to complete multi-tasks dynamic loco-manipulation applications such as efficiently picking up and dropping off objects while turning and walking.

cs.RO↗