arXiv subjects

Mingyang Sun

Publications and source records attributed to Mingyang Sun.

At least 55 records · Page 3Linked to original sources

arXiv2025

VLA-RFT: Vision-Language-Action Reinforcement Fine-tuning with Verified Rewards in World Simulators

Vision-Language-Action (VLA) models enable embodied decision-making but rely heavily on imitation learning, leading to compounding errors and poor robustness under distribution shift. Reinforcement learning (RL) can mitigate these issues yet typically demands costly real-world interactions or suffers from sim-to-real gaps. We introduce VLA-RFT, a reinforcement fine-tuning framework that leverages a data-driven world model as a controllable simulator. Trained from real interaction data, the simulator predicts future visual observations conditioned on actions, allowing policy rollouts with dense, trajectory-level rewards derived from goal-achieving references. This design delivers an efficient and action-aligned learning signal, drastically lowering sample requirements. With fewer than 400 fine-tuning steps, VLA-RFT surpasses strong supervised baselines and achieves greater efficiency than simulator-based RL. Moreover, it exhibits strong robustness under perturbed conditions, sustaining stable task execution. Our results establish world-model-based RFT as a practical post-training paradigm to enhance the generalization and robustness of VLA models. For more details, please refer to https://vla-rft.github.io/.

cs.RO↗

arXiv2025

CARP: Visuomotor Policy Learning via Coarse-to-Fine Autoregressive Prediction

In robotic visuomotor policy learning, diffusion-based models have achieved significant success in improving the accuracy of action trajectory generation compared to traditional autoregressive models. However, they suffer from inefficiency due to multiple denoising steps and limited flexibility from complex constraints. In this paper, we introduce Coarse-to-Fine AutoRegressive Policy (CARP), a novel paradigm for visuomotor policy learning that redefines the autoregressive action generation process as a coarse-to-fine, next-scale approach. CARP decouples action generation into two stages: first, an action autoencoder learns multi-scale representations of the entire action sequence; then, a GPT-style transformer refines the sequence prediction through a coarse-to-fine autoregressive process. This straightforward and intuitive approach produces highly accurate and smooth actions, matching or even surpassing the performance of diffusion-based policies while maintaining efficiency on par with autoregressive policies. We conduct extensive evaluations across diverse settings, including single-task and multi-task scenarios on state-based and image-based simulation benchmarks, as well as real-world tasks. CARP achieves competitive success rates, with up to a 10% improvement, and delivers 10x faster inference compared to state-of-the-art policies, establishing a high-performance, efficient, and flexible paradigm for action generation in robotic tasks.

cs.RO↗

arXiv2025

On-Device Training of PV Power Forecasting Models in a Smart Meter for Grid Edge Intelligence

In this paper, an edge-side model training study is conducted on a resource-limited smart meter. The motivation of grid-edge intelligence and the concept of on-device training are introduced. Then, the technical preparation steps for on-device training are described. A case study on the task of photovoltaic power forecasting is presented, where two representative machine learning models are investigated: a gradient boosting tree model and a recurrent neural network model. To adapt to the resource-limited situation in the smart meter, "mixed"- and "reduced"-precision training schemes are also devised. Experiment results demonstrate the feasibility of economically achieving grid-edge intelligence via the existing advanced metering infrastructures.

cs.LG↗

arXiv2025

AdapThink: Adaptive Thinking Preferences for Reasoning Language Model

Reinforcement Learning (RL)-based post-training has significantly advanced the complex reasoning capabilities of language models, fostering sophisticated self-reflection processes. However, this ``slow thinking'' paradigm presents a critical challenge to reasoning efficiency: models may expend excessive computation on simple questions and shift reasoning prematurely for complex ones. Previous mechanisms typically rely on static length budgets or predefined rules, lacking the adaptability for varying question complexities and models' evolving capabilities. To this end, we propose AdapThink, an adaptive post-training framework designed to induce more efficient thinking while maintaining the performance of reasoning language models. Specifically, AdapThink incorporates two key mechanisms: 1) A group-relative reward function that leverages model confidence and response's characteristic to dynamically adjust the preference of reflection-related transition words without resorting to a fixed length preference. 2) A diversity-aware sampling mechanism that balances the training group's solution accuracy with reasoning diversity via an entropy-guided score. Experiments on several mathematical reasoning datasets with DeepSeek-distilled models demonstrate AdapThink's advantages in enabling adaptive reasoning patterns and mitigating the inefficiencies.

cs.LG↗

arXiv2025

PIG: Physically-based Multi-Material Interaction with 3D Gaussians

3D Gaussian Splatting has achieved remarkable success in reconstructing both static and dynamic 3D scenes. However, in a scene represented by 3D Gaussian primitives, interactions between objects suffer from inaccurate 3D segmentation, imprecise deformation among different materials, and severe rendering artifacts. To address these challenges, we introduce PIG: Physically-Based Multi-Material Interaction with 3D Gaussians, a novel approach that combines 3D object segmentation with the simulation of interacting objects in high precision. Firstly, our method facilitates fast and accurate mapping from 2D pixels to 3D Gaussians, enabling precise 3D object-level segmentation. Secondly, we assign unique physical properties to correspondingly segmented objects within the scene for multi-material coupled interactions. Finally, we have successfully embedded constraint scales into deformation gradients, specifically clamping the scaling and rotation properties of the Gaussian primitives to eliminate artifacts and achieve geometric fidelity and visual consistency. Experimental results demonstrate that our method not only outperforms the state-of-the-art (SOTA) in terms of visual quality, but also opens up new directions and pipelines for the field of physically realistic scene generation.

cs.GR↗

arXiv2025

Think Twice, Act Once: A Co-Evolution Framework of LLM and RL for Large-Scale Decision Making

Recent advancements in Large Language Models (LLMs) and Reinforcement Learning (RL) have shown significant promise in decision-making tasks. Nevertheless, for large-scale industrial decision problems, both approaches face distinct challenges: LLMs lack real-time long-sequence decision-making capabilities, while RL struggles with sample efficiency in vast action spaces. To bridge this gap, we propose Agents Co-Evolution (ACE), a synergistic framework between LLMs and RL agents for large-scale decision-making scenarios. ACE introduces a dual-role trajectory refinement mechanism where LLMs act as both Policy Actor and Value Critic during RL's training: the Actor refines suboptimal actions via multi-step reasoning and environment validation, while the Critic performs temporal credit assignment through trajectory-level reward shaping. Concurrently, RL agent enhances LLMs' task-specific decision-making with high-quality fine-tuning datasets generated via prioritized experience replay. Through extensive experiments across multiple power grid operation challenges with action spaces exceeding 60K discrete actions, ACE demonstrates superior performance over existing RL methods and LLM-based methods.

cs.AI↗

arXiv2025

PSGraph: Differentially Private Streaming Graph Synthesis by Considering Temporal Dynamics

Streaming graphs are ubiquitous in daily life, such as evolving social networks and dynamic communication systems. Due to the sensitive information contained in the graph, directly sharing the streaming graphs poses significant privacy risks. Differential privacy, offering strict theoretical guarantees, has emerged as a standard approach for private graph data synthesis. However, existing methods predominantly focus on static graph publishing, neglecting the intrinsic relationship between adjacent graphs, thereby resulting in limited performance in streaming data publishing scenarios. To address this gap, we propose PSGraph, the first differentially private streaming graph synthesis framework that integrates temporal dynamics. PSGraph adaptively adjusts the privacy budget allocation mechanism by analyzing the variations in the current graph compared to the previous one for conserving the privacy budget. Moreover, PSGraph aggregates information across various timestamps and adopts crucial post-processing techniques to enhance the synthetic streaming graphs. We conduct extensive experiments on four real-world datasets under five commonly used metrics. The experimental results demonstrate the superiority of our proposed PSGraph.

cs.CR↗

arXiv2025

QUART-Online: Latency-Free Large Multimodal Language Model for Quadruped Robot Learning

This paper addresses the inherent inference latency challenges associated with deploying multimodal large language models (MLLM) in quadruped vision-language-action (QUAR-VLA) tasks. Our investigation reveals that conventional parameter reduction techniques ultimately impair the performance of the language foundation model during the action instruction tuning phase, making them unsuitable for this purpose. We introduce a novel latency-free quadruped MLLM model, dubbed QUART-Online, designed to enhance inference efficiency without degrading the performance of the language foundation model. By incorporating Action Chunk Discretization (ACD), we compress the original action representation space, mapping continuous action values onto a smaller set of discrete representative vectors while preserving critical information. Subsequently, we fine-tune the MLLM to integrate vision, language, and compressed actions into a unified semantic space. Experimental results demonstrate that QUART-Online operates in tandem with the existing MLLM system, achieving real-time inference in sync with the underlying controller frequency, significantly boosting the success rate across various tasks by 65%. Our project page is https://quart-online.github.io.

cs.RO↗

arXiv2025

2DGS-Avatar: Animatable High-fidelity Clothed Avatar via 2D Gaussian Splatting

Real-time rendering of high-fidelity and animatable avatars from monocular videos remains a challenging problem in computer vision and graphics. Over the past few years, the Neural Radiance Field (NeRF) has made significant progress in rendering quality but behaves poorly in run-time performance due to the low efficiency of volumetric rendering. Recently, methods based on 3D Gaussian Splatting (3DGS) have shown great potential in fast training and real-time rendering. However, they still suffer from artifacts caused by inaccurate geometry. To address these problems, we propose 2DGS-Avatar, a novel approach based on 2D Gaussian Splatting (2DGS) for modeling animatable clothed avatars with high-fidelity and fast training performance. Given monocular RGB videos as input, our method generates an avatar that can be driven by poses and rendered in real-time. Compared to 3DGS-based methods, our 2DGS-Avatar retains the advantages of fast training and rendering while also capturing detailed, dynamic, and photo-realistic appearances. We conduct abundant experiments on popular datasets such as AvatarRex and THuman4.0, demonstrating impressive performance in both qualitative and quantitative metrics.

cs.CV↗

arXiv2025

SrSv: Integrating Sequential Rollouts with Sequential Value Estimation for Multi-agent Reinforcement Learning

Although multi-agent reinforcement learning (MARL) has shown its success across diverse domains, extending its application to large-scale real-world systems still faces significant challenges. Primarily, the high complexity of real-world environments exacerbates the credit assignment problem, substantially reducing training efficiency. Moreover, the variability of agent populations in large-scale scenarios necessitates scalable decision-making mechanisms. To address these challenges, we propose a novel framework: Sequential rollout with Sequential value estimation (SrSv). This framework aims to capture agent interdependence and provide a scalable solution for cooperative MARL. Specifically, SrSv leverages the autoregressive property of the Transformer model to handle varying populations through sequential action rollout. Furthermore, to capture the interdependence of policy distributions and value functions among multiple agents, we introduce an innovative sequential value estimation methodology and integrates the value approximation into an attention-based sequential model. We evaluate SrSv on three benchmarks: Multi-Agent MuJoCo, StarCraft Multi-Agent Challenge, and DubinsCars. Experimental results demonstrate that SrSv significantly outperforms baseline methods in terms of training efficiency without compromising convergence performance. Moreover, when implemented in a large-scale DubinsCar system with 1,024 agents, our framework surpasses existing benchmarks, highlighting the excellent scalability of SrSv.

cs.AI↗

arXiv2025

SAMG: Offline-to-Online Reinforcement Learning via State-Action-Conditional Offline Model Guidance

Offline-to-online (O2O) reinforcement learning (RL) pre-trains models on offline data and refines policies through online fine-tuning. However, existing O2O RL algorithms typically require maintaining the tedious offline datasets to mitigate the effects of out-of-distribution (OOD) data, which significantly limits their efficiency in exploiting online samples. To address this deficiency, we introduce a new paradigm for O2O RL called State-Action-Conditional Offline \Model Guidance (SAMG). It freezes the pre-trained offline critic to provide compact offline understanding for each state-action sample, thus eliminating the need for retraining on offline data. The frozen offline critic is incorporated with the online target critic weighted by a state-action-adaptive coefficient. This coefficient aims to capture the offline degree of samples at the state-action level, and is updated adaptively during training. In practice, SAMG could be easily integrated with Q-function-based algorithms. Theoretical analysis shows good optimality and lower estimation error. Empirically, SAMG outperforms state-of-the-art O2O RL algorithms on the D4RL benchmark.

cs.LG↗

arXiv2025

Score-Based Diffusion Policy Compatible with Reinforcement Learning via Optimal Transport

Diffusion policies have shown promise in learning complex behaviors from demonstrations, particularly for tasks requiring precise control and long-term planning. However, they face challenges in robustness when encountering distribution shifts. This paper explores improving diffusion-based imitation learning models through online interactions with the environment. We propose OTPR (Optimal Transport-guided score-based diffusion Policy for Reinforcement learning fine-tuning), a novel method that integrates diffusion policies with RL using optimal transport theory. OTPR leverages the Q-function as a transport cost and views the policy as an optimal transport map, enabling efficient and stable fine-tuning. Moreover, we introduce masked optimal transport to guide state-action matching using expert keypoints and a compatibility-based resampling strategy to enhance training stability. Experiments on three simulation tasks demonstrate OTPR's superior performance and robustness compared to existing methods, especially in complex and sparse-reward environments. In sum, OTPR provides an effective framework for combining IL and RL, achieving versatile and reliable policy learning. The code will be released at https://github.com/Sunmmyy/OTPR.git.

cs.LG↗

arXiv2024

Score and Distribution Matching Policy: Advanced Accelerated Visuomotor Policies via Matched Distillation

Visual-motor policy learning has advanced with architectures like diffusion-based policies, known for modeling complex robotic trajectories. However, their prolonged inference times hinder high-frequency control tasks requiring real-time feedback. While consistency distillation (CD) accelerates inference, it introduces errors that compromise action quality. To address these limitations, we propose the Score and Distribution Matching Policy (SDM Policy), which transforms diffusion-based policies into single-step generators through a two-stage optimization process: score matching ensures alignment with true action distributions, and distribution matching minimizes KL divergence for consistency. A dual-teacher mechanism integrates a frozen teacher for stability and an unfrozen teacher for adversarial training, enhancing robustness and alignment with target distributions. Evaluated on a 57-task simulation benchmark, SDM Policy achieves a 6x inference speedup while having state-of-the-art action quality, providing an efficient and reliable framework for high-frequency robotic tasks.

cs.RO↗

arXiv2024

Toward Robust Incomplete Multimodal Sentiment Analysis via Hierarchical Representation Learning

Multimodal Sentiment Analysis (MSA) is an important research area that aims to understand and recognize human sentiment through multiple modalities. The complementary information provided by multimodal fusion promotes better sentiment analysis compared to utilizing only a single modality. Nevertheless, in real-world applications, many unavoidable factors may lead to situations of uncertain modality missing, thus hindering the effectiveness of multimodal modeling and degrading the model's performance. To this end, we propose a Hierarchical Representation Learning Framework (HRLF) for the MSA task under uncertain missing modalities. Specifically, we propose a fine-grained representation factorization module that sufficiently extracts valuable sentiment information by factorizing modality into sentiment-relevant and modality-specific representations through crossmodal translation and sentiment semantic reconstruction. Moreover, a hierarchical mutual information maximization mechanism is introduced to incrementally maximize the mutual information between multi-scale representations to align and reconstruct the high-level semantics in the representations. Ultimately, we propose a hierarchical adversarial learning mechanism that further aligns and adapts the latent distribution of sentiment-relevant representations to produce robust joint multimodal representations. Comprehensive experiments on three datasets demonstrate that HRLF significantly improves MSA performance under uncertain modality missing cases.

cs.CL↗

arXiv2024

PhysHand: A Hand Simulation Model with Physiological Geometry, Physical Deformation, and Accurate Contact Handling

In virtual Hand-Object Interaction (HOI) scenarios, the authenticity of the hand's deformation is important to immersive experience, such as natural manipulation or tactile feedback. Unrealistic deformation arises from simplified hand geometry, neglect of the different physics attributes of the hand, and penetration due to imprecise contact handling. To address these problems, we propose PhysHand, a novel hand simulation model, which enhances the realism of deformation in HOI. First, we construct a physiologically plausible geometry, a layered mesh with a "skin-flesh-skeleton" structure. Second, to satisfy the distinct physics features of different soft tissues, a constraint-based dynamics framework is adopted with carefully designed layer-corresponding constraints to maintain flesh attached and skin smooth. Finally, we employ an SDF-based method to eliminate the penetration caused by contacts and enhance its accuracy by introducing a novel multi-resolution querying strategy. Extensive experiments have been conducted to demonstrate the outstanding performance of PhysHand in calculating deformations and handling contacts. Compared to existing methods, our PhysHand: 1) can compute both physiologically and physically plausible deformation; 2) significantly reduces the depth and count of penetration in HOI.

cs.GR↗

arXiv2024

MaskBEV: Towards A Unified Framework for BEV Detection and Map Segmentation

Accurate and robust multimodal multi-task perception is crucial for modern autonomous driving systems. However, current multimodal perception research follows independent paradigms designed for specific perception tasks, leading to a lack of complementary learning among tasks and decreased performance in multi-task learning (MTL) due to joint training. In this paper, we propose MaskBEV, a masked attention-based MTL paradigm that unifies 3D object detection and bird's eye view (BEV) map segmentation. MaskBEV introduces a task-agnostic Transformer decoder to process these diverse tasks, enabling MTL to be completed in a unified decoder without requiring additional design of specific task heads. To fully exploit the complementary information between BEV map segmentation and 3D object detection tasks in BEV space, we propose spatial modulation and scene-level context aggregation strategies. These strategies consider the inherent dependencies between BEV segmentation and 3D detection, naturally boosting MTL performance. Extensive experiments on nuScenes dataset show that compared with previous state-of-the-art MTL methods, MaskBEV achieves 1.3 NDS improvement in 3D object detection and 2.7 mIoU improvement in BEV map segmentation, while also demonstrating slightly leading inference speed.

cs.CV↗

arXiv2024

HybridOcc: NeRF Enhanced Transformer-based Multi-Camera 3D Occupancy Prediction

Vision-based 3D semantic scene completion (SSC) describes autonomous driving scenes through 3D volume representations. However, the occlusion of invisible voxels by scene surfaces poses challenges to current SSC methods in hallucinating refined 3D geometry. This paper proposes HybridOcc, a hybrid 3D volume query proposal method generated by Transformer framework and NeRF representation and refined in a coarse-to-fine SSC prediction framework. HybridOcc aggregates contextual features through the Transformer paradigm based on hybrid query proposals while combining it with NeRF representation to obtain depth supervision. The Transformer branch contains multiple scales and uses spatial cross-attention for 2D to 3D transformation. The newly designed NeRF branch implicitly infers scene occupancy through volume rendering, including visible and invisible voxels, and explicitly captures scene depth rather than generating RGB color. Furthermore, we present an innovative occupancy-aware ray sampling method to orient the SSC task instead of focusing on the scene surface, further improving the overall performance. Extensive experiments on nuScenes and SemanticKITTI datasets demonstrate the effectiveness of our HybridOcc on the SSC task.

cs.CV↗

arXiv2024

Correlation-Decoupled Knowledge Distillation for Multimodal Sentiment Analysis with Incomplete Modalities

Multimodal sentiment analysis (MSA) aims to understand human sentiment through multimodal data. Most MSA efforts are based on the assumption of modality completeness. However, in real-world applications, some practical factors cause uncertain modality missingness, which drastically degrades the model's performance. To this end, we propose a Correlation-decoupled Knowledge Distillation (CorrKD) framework for the MSA task under uncertain missing modalities. Specifically, we present a sample-level contrastive distillation mechanism that transfers comprehensive knowledge containing cross-sample correlations to reconstruct missing semantics. Moreover, a category-guided prototype distillation mechanism is introduced to capture cross-category correlations using category prototypes to align feature distributions and generate favorable joint representations. Eventually, we design a response-disentangled consistency distillation strategy to optimize the sentiment decision boundaries of the student network through response disentanglement and mutual information maximization. Comprehensive experiments on three datasets indicate that our framework can achieve favorable improvements compared with several baselines.

cs.CV↗