Cong Huang

Bin Yu, Yao Zhang, Haishan Liu et al.

Vision-language-action (VLA) models are built on the premise that semantic understanding from pretrained language or vision-language backbones should guide robot action prediction. Yet robot fine-tuning is optimized as imitation over task-specific action distributions, and many evaluations can be solved through visual or instruction-action shortcuts. We introduce RoboSemanticBench (RSB), an embodied benchmark for diagnosing semantic grounding in action prediction: whether post-trained VLA models can use complex instruction semantics to select and manipulate the correct physical target. In each episode, a robot receives a multiple-choice math or general-knowledge question, observes candidate answer blocks, and must grasp the block corresponding to the correct answer. RSB covers controlled arithmetic, grade-school mathematical understanding, and commonsense or factual understanding under four-choice and ten-choice suites. Across representative VLA models, we find that many policies learn to grasp candidate blocks but select the semantically correct block at near-random or below-random rates after controlling for grasp success, revealing a persistent gap between backbone-level semantic competence and action prediction.

Sirui Liang, Pengfei Cao, Jian Zhao et al.

Parameter-Efficient finetuning (PEFT) enhances model performance on downstream tasks by updating a minimal subset of parameters. Representation finetuning (ReFT) methods further improve efficiency by freezing model weights and optimizing internal representations with fewer parameters than PEFT, outperforming PEFT on several tasks. However, ReFT exhibits a significant performance decline on mathematical reasoning tasks. To address this problem, the paper demonstrates that ReFT's poor performance on mathematical tasks primarily stems from its struggle to generate effective reasoning prefixes during the early inference phase. Moreover, ReFT disturbs the numerical encoding and the error accumulats during the CoT stage. Based on these observations, this paper proposes Bias-REstrained Prefix Representation FineTuning (BREP ReFT), which enhances ReFT's mathematical reasoning capability by truncating training data to optimize the generation of initial reasoning prefixes, intervening on the early inference stage to prevent error accumulation, and constraining the intervention vectors' magnitude to avoid disturbing numerical encoding. Extensive experiments across diverse model architectures demonstrate BREP's superior effectiveness, efficiency, and robust generalization capability, outperforming both standard ReFT and weight-based PEFT methods on the task of mathematical reasoning. The source code is available at https://github.com/LiangThree/BREP.

Cong Huang, Jiahao Li, Bin Li et al.

How to efficiently utilize the temporal features is crucial, yet challenging, for video restoration. The temporal features usually contain various noisy and uncorrelated information, and they may interfere with the restoration of the current frame. This paper proposes learning noise-robust feature representations to help video restoration. We are inspired by that the neural codec is a natural denoiser. In neural codec, the noisy and uncorrelated contents which are hard to predict but cost lots of bits are more inclined to be discarded for bitrate saving. Therefore, we design a neural compression module to filter the noise and keep the most useful information in features for video restoration. To achieve robustness to noise, our compression module adopts a spatial channel-wise quantization mechanism to adaptively determine the quantization step size for each position in the latent. Experiments show that our method can significantly boost the performance on video denoising, where we obtain 0.13 dB improvement over BasicVSR++ with only 0.23x FLOPs. Meanwhile, our method also obtains SOTA results on video deraining and dehazing.

Peng Ren, Haoyang Ge, Chuan Qi et al.

Robots are increasingly expected to execute open ended natural language requests in human environments, which demands reliable long horizon execution under partial observability. This is especially challenging for humanoids because locomotion and manipulation are tightly coupled through stance, reachability, and balance. We present a humanoid agent framework that turns VLM plans into verifiable task programs and closes the loop with multi object 3D geometric supervision. A VLM planner compiles each instruction into a typed JSON sequence of subtasks with explicit predicate based preconditions and success conditions. Using SAM3 and RGB-D, we ground all task relevant entities in 3D, estimate object centroids and extents, and evaluate predicates over stable frames to obtain condition level diagnostics. The supervisor uses these diagnostics to verify subtask completion and to provide condition-level feedback for progression and replanning. We execute each subtask by coordinating humanoid locomotion and whole-body manipulation, selecting feasible motion primitives under reachability and balance constraints. Experiments on tabletop manipulation and long horizon humanoid loco manipulation tasks show improved robustness from multi object grounding, temporal stability, and recovery driven replanning.

Changti Wu, Jiahuai Mao, Yuzhuo Miao et al.

Large-scale Visual Instruction Tuning (VIT) has become a key paradigm for advancing the performance of vision-language models (VLMs) across various multimodal tasks. However, training on the large-scale datasets is computationally expensive and inefficient due to redundancy in the data, which motivates the need for multimodal data selection to improve training efficiency. Existing data selection methods for VIT either require costly training or gradient computation. Training-free alternatives often depend on proxy models or datasets, instruction-agnostic representations, and pairwise similarity with quadratic complexity, limiting scalability and representation fidelity. In this work, we propose ScalSelect, a scalable training-free multimodal data selection method with linear-time complexity with respect to the number of samples, eliminating the need for external models or auxiliary datasets. ScalSelect first constructs sample representations by extracting visual features most attended by instruction tokens in the target VLM, capturing instruction-relevant information. It then identifies samples whose representations best approximate the dominant subspace of the full dataset representations, enabling scalable importance scoring without pairwise comparisons. Extensive experiments across multiple VLMs, datasets, and selection budgets demonstrate that ScalSelect achieves over 97.5% of the performance of training on the full dataset using only 16% of the data, and even outperforms full-data training in some settings. The code is available at \href{https://github.com/ChangtiWu/ScalSelect}{ScalSelect}.

Bin Yu, Shijie Lian, Xiaopeng Lin et al.

Vision-Language-Action (VLA) models leverage Multimodal Large Language Models (MLLMs) for robotic control, but recent studies reveal that MLLMs exhibit limited spatial intelligence due to training predominantly on 2D data, resulting in inadequate 3D perception for manipulation tasks. While recent approaches incorporate specialized 3D vision models such as VGGT to enhance spatial understanding, they employ diverse integration mechanisms without systematic investigation, leaving the optimal fusion strategy unclear. We conduct a comprehensive pilot study comparing nine VGGT integration schemes on standardized benchmarks and find that semantic-conditioned gated fusion, which adaptively balances 2D semantic and 3D geometric features based on task context, achieved the strongest performance among all nine evaluated fusion schemes in our pilot study. We present 3D-Mix, a plug-and-play module that integrates into diverse VLA architectures (GR00T-style and $π$-style) without modifying existing MLLM or action expert components. Experiments across six MLLM series (nine model variants, 2B--8B parameters) on SIMPLER and LIBERO show that 3D-Mix delivers consistent performance gains, averaging +7.0% on the out-of-domain (OOD) SIMPLER benchmark across all nine GR00T-style variants, establishing a principled approach for enhancing spatial intelligence in VLA systems.

Bin Yu, Shijie Lian, Xiaopeng Lin et al.

Standard Vision-Language-Action (VLA) models typically fine-tune a monolithic Vision-Language Model (VLM) backbone explicitly for robotic control. However, this approach creates a critical tension between maintaining high-level general semantic understanding and learning low-level, fine-grained sensorimotor skills, often leading to "catastrophic forgetting" of the model's open-world capabilities. To resolve this conflict, we introduce TwinBrainVLA, a novel architecture that coordinates a generalist VLM retaining universal semantic understanding and a specialist VLM dedicated to embodied proprioception for joint robotic control. TwinBrainVLA synergizes a frozen "Left Brain", which retains robust general visual reasoning, with a trainable "Right Brain", specialized for embodied perception, via a novel Asymmetric Mixture-of-Transformers (AsyMoT) mechanism. This design allows the Right Brain to dynamically query semantic knowledge from the frozen Left Brain and fuse it with proprioceptive states, providing rich conditioning for a Flow-Matching Action Expert to generate precise continuous controls. Extensive experiments on SimplerEnv and RoboCasa benchmarks demonstrate that TwinBrainVLA achieves superior manipulation performance compared to state-of-the-art baselines while explicitly preserving the comprehensive visual understanding capabilities of the pre-trained VLM, offering a promising direction for building general-purpose robots that simultaneously achieve high-level semantic understanding and low-level physical dexterity.

Shijie Lian, Bin Yu, Xiaopeng Lin et al.

Vision-language-action models have advanced rapidly, but robot trajectories alone provide limited coverage for learning broad physical understanding. PhysBrain 1.0 studies a complementary route: converting large-scale human egocentric video into structured physical commonsense supervision before robot adaptation. Our data engine extracts scene elements, spatial dynamics, action execution, and depth-aware relations, then turns them into question-answer supervision for training PhysBrain VLMs. The resulting physical priors are further transferred to VLA policies through a capability-preserving and language-sensitive adaptation design. Across multimodal QA benchmarks and embodied control benchmarks, including ERQA, PhysBench, SimplerEnv-WidowX, LIBERO, and RoboCasa, PhysBrain 1.0 achieves SOTA results and shows especially strong out-of-domain performance on SimplerEnv. These results suggest that scaling physical commonsense from human interaction video can provide an effective bridge from multimodal understanding to robot action.

Shijie Lian, Bin Yu, Xiaopeng Lin et al.

Robot imitation data are often multimodal: similar visual-language observations may be followed by different action chunks because human demonstrators act with different short-horizon intents, task phases, or recent context. Existing frame-conditioned VLA policies infer each chunk from the current observation and instruction alone, so under partial observability they may resample different intents across adjacent replanning steps, leading to inter-chunk conflict and unstable execution. We introduce IntentVLA, a history-conditioned VLA framework that encodes recent visual observations into a compact short-horizon intent representation and uses it to condition chunk generation. We further introduce AliasBench, a 12-task ambiguity-aware benchmark on RoboTwin2 with matched training data and evaluation environments that isolate short-horizon observation aliasing. Across AliasBench, SimplerEnv, LIBERO, and RoboCasa, IntentVLA improves rollout stability and outperforms strong VLA baselines

Bin Yu, Shijie Lian, Xiaopeng Lin et al.

Vision-Language-Action (VLA) policies are commonly trained from dense robot demonstration trajectories, often collected through teleoperation, by sampling every recorded frame as if it provided equally useful supervision. We argue that this convention creates a temporal supervision imbalance: long low-change segments dominate the training stream, while manipulation-critical transitions such as alignment, contact, grasping, and release appear only sparsely. We introduce FrameSkip, a data-layer frame selection framework that scores trajectory frames using action variation, visual-action coherence, task-progress priors, and gripper-transition preservation, then remaps training samples toward high-importance frames under a target retention ratio. Because FrameSkip operates only in the dataloader, it leaves the VLA architecture, action head, training objective, and inference procedure unchanged. Across RoboCasa-GR1, SimplerEnv, and LIBERO, FrameSkip improves the success-retention trade-off over full-frame training and simpler frame selection variants, achieving a macro-average success rate of 76.15% across the three benchmarks compared with 66.50% for full-frame training while using a compressed trajectory view that retains 20% of unique frames in the main setting.

Hang Yuan, Xiaolin Hu, Yan Wan et al.

Text-driven controllable dance generation remains under-explored, primarily due to the severe scarcity of high-quality datasets and the inherent difficulty of articulating complex choreographies. Characterizing dance is particularly challenging owing to its intricate spatial dynamics, strong directionality, and the highly decoupled movements of distinct body parts. To overcome these bottlenecks, we bridge principles from dance studies, human anatomy, and biomechanics to propose \textit{Choreographic Syntax}, a novel theoretical framework with a tailored annotation system. Grounded in this syntax, we combine professional dance archives with high-fidelity motion capture data to construct \textbf{DanceFlow}, the most fine-grained dance dataset to date. It encompasses 41 hours of high-quality motions paired with 6.34 million words of detailed descriptions. At the model level, we introduce \textbf{DanceCrafter}, a tailored motion transformer built upon the Momentum Human Rig. To circumvent optimization instabilities, we construct a continuous manifold motion representation paired with a hybrid normalization strategy. Furthermore, we design an anatomy-aware loss to explicitly regulate the decoupled nature of body parts. Together, these adaptations empower DanceCrafter to achieve the high-fidelity and stable generation of complex dance sequences. Extensive evaluations and user studies demonstrate our state-of-the-art performance in motion quality, fine-grained controllability, and generation naturalness.

Shijie Lian, Bin Yu, Xiaopeng Lin et al.

Vision-Language-Action (VLA) models have shown promise in robot manipulation but often struggle to generalize to new instructions or complex multi-task scenarios. We identify a critical pathology in current training paradigms where goal-driven data collection creates a dataset bias. In such datasets, language instructions are highly predictable from visual observations alone, causing the conditional mutual information between instructions and actions to vanish, a phenomenon we term Information Collapse. Consequently, models degenerate into vision-only policies that ignore language constraints and fail in out-of-distribution (OOD) settings. To address this, we propose BayesianVLA, a novel framework that enforces instruction following via Bayesian decomposition. By introducing learnable Latent Action Queries, we construct a dual-branch architecture to estimate both a vision-only prior $p(a \mid v)$ and a language-conditioned posterior $π(a \mid v, \ell)$. We then optimize the policy to maximize the conditional Pointwise Mutual Information (PMI) between actions and instructions. This objective effectively penalizes the vision shortcut and rewards actions that explicitly explain the language command. Without requiring new data, BayesianVLA significantly improves generalization. Extensive experiments across on SimplerEnv and RoboCasa demonstrate substantial gains, including an 11.3% improvement on the challenging OOD SimplerEnv benchmark, validating the ability of our approach to robustly ground language in action.

Zhaoyang Yang, Yurun Jin, Lizhe Qi et al.

Pretrained video diffusion models provide powerful spatiotemporal generative priors, making them a natural foundation for robotic world models. While recent world-action models jointly optimize future videos and actions, they predominantly treat video generation as an auxiliary representation for policy learning. Consequently, they insufficiently explore the inverse problem: leveraging action signals to guide video synthesis, thereby often failing to preserve precise robot spatial geometry and fine-grained robot-object interaction dynamics in the generated rollouts. To bridge this gap, we present EA-WM, an Event-Aware Generative World Model that effectively closes the loop between kinematic control and visual perception. Rather than injecting joint or end-effector actions as abstract, low-dimensional tokens, EA-WM projects actions and kinematic states directly into the target camera view as Structured Kinematic-to-Visual Action Fields. To fully exploit this geometrically grounded representation, we introduce event-aware bidirectional fusion blocks that modulate cross-branch attention, capturing object state changes and interaction dynamics. Evaluated on the comprehensive WorldArena benchmark, EA-WM achieves state-of-the-art performance, outperforming existing baselines by a significant margin.

Yuxuan Tian, Yurun Jin, Bin Yu et al.

Robotic manipulation critically requires reasoning about future spatial-temporal interactions, yet existing VLA policies and world-model-enhanced policies do not fully model action-relevant spatial-temporal interaction structure. We propose STARRY, a world-model-enhanced action-generation policy that aligns spatial-temporal prediction with action generation. STARRY jointly denoises future spatial-temporal latents and action sequences, and introduces Geometry-Aware Selective Attention Modulation to convert predicted depth and end-effector geometry into token-aligned weights for selective action-attention modulation. On RoboTwin 2.0, STARRY achieves 93.82% / 93.30% average success under Clean and Randomized settings. Real-world experiments further improve average success from 42.5% to 70.8% over $π_{0.5}$, demonstrating the effectiveness of action-centric spatial-temporal world modeling for spatial-temporally demanding robotic action generation.

Peng Ren, Chuan Qi, Haoyang Ge et al.

Dynamic ball-interaction tasks remain challenging for robots because they require tight perception-action coupling under limited reaction time. This challenge is especially pronounced in humanoid racket sports, where successful interception depends on accurate visual tracking, trajectory prediction, coordinated stepping, and stable whole-body striking. Existing robotic racket-sport systems often rely on external motion capture for state estimation or on task-specific low-level controllers that must be retrained across tasks and platforms. We present CyboRacket, a hierarchical perception-to-action framework for humanoid racket sports that integrates onboard visual perception, physics-based trajectory prediction, and large-scale pre-trained whole-body control. The framework uses onboard cameras to track the incoming object, predicts its future trajectory, and converts the estimated interception state into target end-effector and base-motion commands for whole-body execution by SONIC on the Unitree G1 humanoid robot. We evaluate the proposed framework in a vision-based humanoid tennis-hitting task. Experimental results demonstrate real-time visual tracking, trajectory prediction, and successful striking using purely onboard sensing.

Xiaolin Hu, Hang Yuan, Xinzhu Sang et al.

Developing expressive and responsive conversational digital humans is a cornerstone of next-generation human-computer interaction. While large language models (LLMs) have significantly enhanced dialogue capabilities, most current systems still rely on cascaded architectures that connect independent modules. These pipelines are often plagued by accumulated errors, high latency, and poor real-time performance. Lacking access to the underlying conversational context, these pipelines inherently prioritize rigid lip-sync over emotional depth. To address these challenges, we propose A$^2$-LLM, an end-to-end conversational audio avatar large language model that jointly reasons about language, audio prosody, and 3D facial motion within a unified framework. To facilitate training, we introduce FLAME-QA, a high-quality multimodal dataset designed to align semantic intent with expressive facial dynamics within a QA format. By leveraging deep semantic understanding, A$^2$-LLM generates emotionally rich facial movements beyond simple lip-synchronization. Experimental results demonstrate that our system achieves superior emotional expressiveness while maintaining real-time efficiency (500 ms latency, 0.7 RTF).

Xingxi Yin, Yicheng Li, Gong Yan et al.

Sequential identity consistency under precise transient attribute control remains a long-standing challenge in controllable visual storytelling. Existing datasets lack sufficient fidelity and fail to disentangle stable identities from transient attributes, limiting structured control over pose, expression, and scene composition and thus constraining reliable sequential synthesis. To address this gap, we introduce \textbf{2K-Characters-10K-Stories}, a multi-modal stylized narrative dataset of \textbf{2{,}000} uniquely stylized characters appearing across \textbf{10{,}000} illustration stories. It is the first dataset that pairs large-scale unique identities with explicit, decoupled control signals for sequential identity consistency. We introduce a \textbf{Human-in-the-Loop pipeline (HiL)} that leverages expert-verified character templates and LLM-guided narrative planning to generate highly-aligned structured data. A \textbf{decoupled control} scheme separates persistent identity from transient attributes -- pose and expression -- while a \textbf{Quality-Gated loop} integrating MMLM evaluation, Auto-Prompt Tuning, and Local Image Editing enforces pixel-level consistency. Extensive experiments demonstrate that models fine-tuned on our dataset achieves performance comparable to closed-source models in generating visual narratives.

Tao Hu, Jizheng Xu, Cong Huang et al.

For fine-grained visual classification, objects usually share similar geometric structure but present variant local appearance and different pose. Therefore, localizing and extracting discriminative local features play a crucial role in accurate category prediction. Existing works either pay attention to limited object parts or train isolated networks for locating and classification. In this paper, we propose Weakly Supervised Bilinear Attention Network (WS-BAN) to solve these issues. It jointly generates a set of attention maps (region-of-interest maps) to indicate the locations of object's parts and extracts sequential part features by Bilinear Attention Pooling (BAP). Besides, we propose attention regularization and attention dropout to weakly supervise the generating process of attention maps. WS-BAN can be trained end-to-end and achieves the state-of-the-art performance on multiple fine-grained classification datasets, including CUB-200-2011, Stanford Car and FGVC-Aircraft, which demonstrated its effectiveness.