Sipeng Zheng

Sipeng Zheng, Qi Zhang, Bei Liu et al.

In this paper we provide the technique report of Ego4D natural language query challenge in CVPR 2022. Natural language query task is challenging due to the requirement of comprehensive understanding of video contents. Most previous works address this task based on third-person view datasets while few research interest has been placed in the ego-centric view by far. Great progress has been made though, we notice that previous works can not adapt well to ego-centric view datasets e.g., Ego4D mainly because of two reasons: 1) most queries in Ego4D have a excessively small temporal duration (e.g., less than 5 seconds); 2) queries in Ego4D are faced with much more complex video understanding of long-term temporal orders. Considering these, we propose our solution of this challenge to solve the above issues.

Qi Zhang, Sipeng Zheng, Qin Jin

Temporal video grounding (TVG) aims to retrieve the time interval of a language query from an untrimmed video. A significant challenge in TVG is the low "Semantic Noise Ratio (SNR)", which results in worse performance with lower SNR. Prior works have addressed this challenge using sophisticated techniques. In this paper, we propose a no-frills TVG model that consists of two core modules, namely multi-scale neighboring attention and zoom-in boundary detection. The multi-scale neighboring attention restricts each video token to only aggregate visual contexts from its neighbor, enabling the extraction of the most distinguishing information with multi-scale feature hierarchies from high-ratio noises. The zoom-in boundary detection then focuses on local-wise discrimination of the selected top candidates for fine-grained grounding adjustment. With an end-to-end training strategy, our model achieves competitive performance on different TVG benchmarks, while also having the advantage of faster inference speed and lighter model parameters, thanks to its lightweight architecture.

Bin Cao, Sipeng Zheng, Hao Luo et al.

Text-to-motion (T2M) generation aims to create realistic human movements from text descriptions, with promising applications in animation and robotics. Despite recent progress, current T2M models perform poorly on unseen text descriptions due to the small scale and limited diversity of existing motion datasets. To address this problem, we introduce OpenT2M, a million-level, high-quality, and open-source motion dataset containing over 2800 hours of human motion. Each sequence undergoes rigorous quality control through physical feasibility validation and multi-granularity filtering, with detailed second-wise text annotations. We also develop an automated pipeline for creating long-horizon sequences, enabling complex motion generation. Building upon OpenT2M, we introduce MonoFrill, a pretrained motion model that achieves compelling T2M results without complicated designs or technique tricks as "frills". Its core component is 2D-PRQ, a novel motion tokenizer that captures spatiotemporal dependencies by dividing the human body into biology parts. Experiments show that OpenT2M significantly improves generalization of existing T2M models, while 2D-PRQ achieves superior reconstruction and strong zero-shot performance. We expect OpenT2M and MonoFrill will advance the T2M field by addressing longstanding data quality and benchmarking challenges.

Ludan Ruan, Anwen Hu, Yuqing Song et al.

Multimodal processing has attracted much attention lately especially with the success of pre-training. However, the exploration has mainly focused on vision-language pre-training, as introducing more modalities can greatly complicate model design and optimization. In this paper, we extend the stateof-the-art Vision-Language model CLIP to accommodate the audio modality for Vision-Language-Audio multimodal processing. Specifically, we apply inter-modal and intra-modal contrastive learning to explore the correlation between audio and other modalities in addition to the inner characteristics of the audio modality. Moreover, we further design an audio type token to dynamically learn different audio information type for different scenarios, as both verbal and nonverbal heterogeneous information is conveyed in general audios. Our proposed CLIP4VLA model is validated in different downstream tasks including video retrieval and video captioning, and achieves the state-of-the-art performance on the benchmark datasets of MSR-VTT, VATEX, and Audiocaps.

Sipeng Zheng, Jiazheng Liu, Yicheng Feng et al.

Recent studies have presented compelling evidence that large language models (LLMs) can equip embodied agents with the self-driven capability to interact with the world, which marks an initial step toward versatile robotics. However, these efforts tend to overlook the visual richness of open worlds, rendering the entire interactive process akin to "a blindfolded text-based game." Consequently, LLM-based agents frequently encounter challenges in intuitively comprehending their surroundings and producing responses that are easy to understand. In this paper, we propose Steve-Eye, an end-to-end trained large multimodal model designed to address this limitation. Steve-Eye integrates the LLM with a visual encoder which enables it to process visual-text inputs and generate multimodal feedback. In addition, we use a semi-automatic strategy to collect an extensive dataset comprising 850K open-world instruction pairs, empowering our model to encompass three essential functions for an agent: multimodal perception, foundational knowledge base, and skill prediction and planning. Lastly, we develop three open-world evaluation benchmarks, then carry out extensive experiments from a wide range of perspectives to validate our model's capability to strategically act and plan. Codes and datasets will be released.

Yicheng Feng, Yuxuan Wang, Jiazheng Liu et al.

Recently, various studies have leveraged Large Language Models (LLMs) to help decision-making and planning in environments, and try to align the LLMs' knowledge with the world conditions. Nonetheless, the capacity of LLMs to continuously acquire environmental knowledge and adapt in an open world remains uncertain. In this paper, we propose an approach to spur LLMs to explore the open world, gather experiences, and learn to improve their task-solving capabilities. In this approach, a multi-round feedback-revision mechanism is utilized to encourage LLMs to actively select appropriate revision actions guided by feedback information from the environment. This facilitates exploration and enhances the model's performance. Besides, we integrate sub-task relabeling to assist LLMs in maintaining consistency in sub-task planning and help the model learn the combinatorial nature between tasks, enabling it to complete a wider range of tasks through training based on the acquired exploration experiences. By evaluation in Minecraft, an open-ended sandbox world, we demonstrate that our approach LLaMA-Rider enhances the efficiency of the LLM in exploring the environment, and effectively improves the LLM's ability to accomplish more tasks through fine-tuning with merely 1.3k instances of collected data, showing minimal training costs compared to the baseline using reinforcement learning.

Haiweng Xu, Sipeng Zheng, Hao Luo et al.

Recent Vision-Language-Action (VLA) models report impressive success rates on standard robotic benchmarks, fueling optimism about general-purpose physical intelligence. However, recent evidence suggests a systematic misalignment between standard benchmark success and true embodied reasoning, raising the question of whether these high scores reflect genuine cognitive capability. To address this gap, we introduce BeTTER, a diagnostic Benchmark for Testing True Embodied Reasoning in robotic policies. BeTTER applies targeted causal interventions (e.g., spatial layout shifts, temporal extrapolation) while enforcing kinematic isolation to explicitly decouple high-level reasoning failures from low-level execution limits. Through systematic evaluation, we reveal that state-of-the-art VLAs catastrophically fail in dynamic scenarios, exhibiting severe lexical-kinematic shortcuts, behavioral inertia, and semantic feature collapse. Crucially, our mechanistic analysis traces these symptoms to fundamental architectural bottlenecks - such as capacity compression and myopic downsampling - which systematically degrade the model's foundational semantic representation. We demonstrate that highly static evaluation protocols effectively mask this degradation by allowing optimization to overfit to sensorimotor priors. Supported by real-world robotic validation, our findings confirm that this representational breakdown is not a simulation artifact, highlighting the critical need for future VLA paradigms to resolve the structural tension between high-frequency control and high-level reasoning.

Boshen Xu, Sipeng Zheng, Qin Jin

We humans are good at translating third-person observations of hand-object interactions (HOI) into an egocentric view. However, current methods struggle to replicate this ability of view adaptation from third-person to first-person. Although some approaches attempt to learn view-agnostic representation from large-scale video datasets, they ignore the relationships among multiple third-person views. To this end, we propose a Prompt-Oriented View-agnostic learning (POV) framework in this paper, which enables this view adaptation with few egocentric videos. Specifically, We introduce interactive masking prompts at the frame level to capture fine-grained action information, and view-aware prompts at the token level to learn view-agnostic representation. To verify our method, we establish two benchmarks for transferring from multiple third-person views to the egocentric view. Our extensive experiments on these benchmarks demonstrate the efficiency and effectiveness of our POV framework and prompt tuning techniques in terms of view adaptation and view generalization. Our code is available at \url{https://github.com/xuboshen/pov_acmmm2023}.

Wanpeng Zhang, Hao Luo, Sipeng Zheng et al.

Offline post-training adapts a pretrained robot policy to a target dataset by supervised regression on recorded actions. In practice, robot datasets are heterogeneous: they mix embodiments, camera setups, and demonstrations of varying quality, so many trajectories reflect recovery behavior, inconsistent operator skill, or weakly informative supervision. Uniform post-training gives equal credit to all samples and can therefore average over conflicting or low-attribution data. We propose Posterior-Transition Reweighting (PTR), a reward-free and conservative post-training method that decides how much each training sample should influence the supervised update. For each sample, PTR encodes the observed post-action consequence as a latent target, inserts it into a candidate pool of mismatched targets, and uses a separate transition scorer to estimate a softmax identification posterior over target indices. The posterior-to-uniform ratio defines the PTR score, which is converted into a clipped-and-mixed weight and applied to the original action objective through self-normalized weighted regression. This construction requires no tractable policy likelihood and is compatible with both diffusion and flow-matching action heads. Rather than uniformly trusting all recorded supervision, PTR reallocates credit according to how attributable each sample's post-action consequence is under the current representation, improving conservative offline adaptation to heterogeneous robot data.

Boshen Xu, Sipeng Zheng, Qin Jin

We introduce SPAFormer, an innovative model designed to overcome the combinatorial explosion challenge in the 3D Part Assembly (3D-PA) task. This task requires accurate prediction of each part's poses in sequential steps. As the number of parts increases, the possible assembly combinations increase exponentially, leading to a combinatorial explosion that severely hinders the efficacy of 3D-PA. SPAFormer addresses this problem by leveraging weak constraints from assembly sequences, effectively reducing the solution space's complexity. Since the sequence of parts conveys construction rules similar to sentences structured through words, our model explores both parallel and autoregressive generation. We further strengthen SPAFormer through knowledge enhancement strategies that utilize the attributes of parts and their sequence information, enabling it to capture the inherent assembly pattern and relationships among sequentially ordered parts. We also construct a more challenging benchmark named PartNet-Assembly covering 21 varied categories to more comprehensively validate the effectiveness of SPAFormer. Extensive experiments demonstrate the superior generalization capabilities of SPAFormer, particularly with multi-tasking and in scenarios requiring long-horizon assembly. Code is available at https://github.com/xuboshen/SPAFormer.

Boshen Xu, Yuting Mei, Xinbi Liu et al.

Egocentric video-language pretraining has significantly advanced video representation learning. Humans perceive and interact with a fully 3D world, developing spatial awareness that extends beyond text-based understanding. However, most previous works learn from 1D text or 2D visual cues, such as bounding boxes, which inherently lack 3D understanding. To bridge this gap, we introduce EgoDTM, an Egocentric Depth- and Text-aware Model, jointly trained through large-scale 3D-aware video pretraining and video-text contrastive learning. EgoDTM incorporates a lightweight 3D-aware decoder to efficiently learn 3D-awareness from pseudo depth maps generated by depth estimation models. To further facilitate 3D-aware video pretraining, we enrich the original brief captions with hand-object visual cues by organically combining several foundation models. Extensive experiments demonstrate EgoDTM's superior performance across diverse downstream tasks, highlighting its superior 3D-aware visual understanding. Our code will be released at https://github.com/xuboshen/EgoDTM.

Hao Luo, Wanpeng Zhang, Yicheng Feng et al.

Visual-Language-Action models (VLAs) have advanced generalist robot control by mapping multimodal observations and language instructions directly to actions, but sparse action supervision often encourages shortcut mappings rather than representations of dynamics, contact, and task progress. Recent world-action models introduce future prediction through video rollouts, yet pixel-space prediction is a costly and indirect substrate for control, as it may model visual details irrelevant to action generation and introduces substantial training or inference overhead. We present Being-H0.7, a latent world-action model that brings future-aware reasoning into VLA-style policies without generating future frames. Being-H0.7 inserts learnable latent queries between perception and action as a compact reasoning interface, and trains them with a future-informed dual-branch design: a deployable prior branch infers latent states from the current context, while a training-only posterior branch replaces the queries with embeddings from future observations. Jointly aligning the two branches at the latent reasoning space leads the prior branch to reason future-aware, action-useful structure from current observations alone. At inference, Being-H0.7 discards the posterior branch and performs no visual rollout. Experiments across six simulation benchmarks and diverse real-world tasks show that Being-H0.7 achieves state-of-the-art or comparable performance, combining the predictive benefits of world models with the efficiency and deployability of direct VLA policies.

Hao Luo, Yicheng Feng, Wanpeng Zhang et al.

We introduce Being-H0, a dexterous Vision-Language-Action model (VLA) trained on large-scale human videos. Existing VLAs struggle with complex manipulation tasks requiring high dexterity and generalize poorly to novel scenarios and tasks, primarily due to their reliance on synthetic data with significant sim-to-real gaps or teleoperated demonstrations lacking scale and diversity. To address this data bottleneck, we propose leveraging human hands as a foundation manipulator, capitalizing on the rich dexterity and scalability present in web data. Our approach centers on physical instruction tuning, a novel training paradigm that combines large-scale VLA pretraining from human videos, physical space alignment for 3D reasoning, and post-training adaptation for robotic tasks. Additionally, we introduce a part-level motion tokenization method which achieves millimeter-level reconstruction accuracy to model precise hand trajectories for action learning. To support our proposed paradigm, we further develop a comprehensive data curation pipeline that integrates heterogeneous sources -- including motion capture, VR, and RGB-only videos -- into a large-scale dataset with millions of motion-based instructional instances. We empirically show the excellence of Being-H0 in hand motion generation and instruction following, and it also scales well with model and data sizes. Importantly, we observe the expected gains of Being-H0 in real-world robotic manipulation as physical instruction tuning is applied. More details are available at https://beingbeyond.github.io/Being-H0.

Yicheng Feng, Yijiang Li, Wanpeng Zhang et al.

We present VideoOrion, a Video Large Language Model (Video-LLM) that explicitly captures the key semantic information in videos - the spatial-temporal dynamics of objects throughout the videos. VideoOrion employs expert vision models to extract object dynamics through a detect-segment-track pipeline, encoding them into a set of object tokens by aggregating spatial-temporal object features. Our method addresses the persistent challenge in Video-LLMs of efficiently compressing high-dimensional video data into semantic tokens that are comprehensible to LLMs. Compared to prior methods which resort to downsampling the original video or aggregating visual tokens using resamplers, leading to information loss and entangled semantics, VideoOrion not only offers a more natural and efficient way to derive compact, disentangled semantic representations but also enables explicit object modeling of video content with minimal computational cost. Moreover, the introduced object tokens naturally allow VideoOrion to accomplish video-based referring tasks. Experimental results show that VideoOrion can learn to make good use of the object tokens, and achieves competitive results on both general video question answering and video-based referring benchmarks.

Wanpeng Zhang, Yicheng Feng, Hao Luo et al.

Multimodal large language models (MLLMs) have made significant progress in vision-language understanding, yet effectively aligning different modalities remains a fundamental challenge. We present a framework that unifies multimodal understanding by applying byte-pair encoding to visual tokens. Unlike conventional approaches that rely on modality-specific encoders, our method directly incorporates structural information into visual tokens, mirroring successful tokenization strategies in text-only language models. We introduce a priority-guided encoding scheme that considers both frequency and spatial consistency, coupled with a multi-stage training procedure based on curriculum-driven data composition. These enhancements enable the transformer model to better capture cross-modal relationships and reason with visual information. Comprehensive experiments demonstrate improved performance across diverse vision-language tasks. By bridging the gap between visual and textual representations, our approach contributes to the advancement of more capable and efficient multimodal foundation models.

Junpeng Yue, Zepeng Wang, Yuxuan Wang et al.

This paper focuses on a critical challenge in robotics: translating text-driven human motions into executable actions for humanoid robots, enabling efficient and cost-effective learning of new behaviors. While existing text-to-motion generation methods achieve semantic alignment between language and motion, they often produce kinematically or physically infeasible motions unsuitable for real-world deployment. To bridge this sim-to-real gap, we propose Reinforcement Learning from Physical Feedback (RLPF), a novel framework that integrates physics-aware motion evaluation with text-conditioned motion generation. RLPF employs a motion tracking policy to assess feasibility in a physics simulator, generating rewards for fine-tuning the motion generator. Furthermore, RLPF introduces an alignment verification module to preserve semantic fidelity to text instructions. This joint optimization ensures both physical plausibility and instruction alignment. Extensive experiments show that RLPF greatly outperforms baseline methods in generating physically feasible motions while maintaining semantic correspondence with text instruction, enabling successful deployment on real humanoid robots.

Jiazheng Liu, Sipeng Zheng, Börje F. Karlsson et al.

Multimodal large language models (MLLMs), built on large-scale pre-trained vision towers and language models, have shown great capabilities in multimodal understanding. However, most existing MLLMs are trained on single-turn vision question-answering tasks, which do not accurately reflect real-world human conversations. In this paper, we introduce MMDiag, a multi-turn multimodal dialogue dataset. This dataset is collaboratively generated through deliberately designed rules and GPT assistance, featuring strong correlations between questions, between questions and images, and among different image regions; thus aligning more closely with real-world scenarios. MMDiag serves as a strong benchmark for multi-turn multimodal dialogue learning and brings more challenges to the grounding and reasoning capabilities of MLLMs. Further, inspired by human vision processing, we present DiagNote, an MLLM equipped with multimodal grounding and reasoning capabilities. DiagNote consists of two modules (Deliberate and Gaze) interacting with each other to perform Chain-of-Thought and annotations respectively, throughout multi-turn dialogues. We empirically demonstrate the advantages of DiagNote in both grounding and jointly processing and reasoning with vision and language information over existing MLLMs.

Boyuan Li, Sipeng Zheng, Bin Cao et al.

Extracting human motion from large-scale web videos offers a scalable solution to the data scarcity issue in character animation. However, some human parts in many video frames cannot be seen due to off-screen captures or occlusions. It brings a dilemma: discarding the data missing any part limits scale and diversity, while retaining it compromises data quality and model performance. To address this problem, we propose leveraging credible part-level data extracted from videos to enhance motion generation via a robust part-aware masked autoregression model. First, we decompose a human body into five parts and detect the parts clearly seen in a video frame as "credible". Second, the credible parts are encoded into latent tokens by our proposed part-aware variational autoencoder. Third, we propose a robust part-level masked generation model to predict masked credible parts, while ignoring those noisy parts. In addition, we contribute K700-M, a challenging new benchmark comprising approximately 200k real-world motion sequences, for evaluation. Experimental results indicate that our method successfully outperforms baselines on both clean and noisy datasets in terms of motion quality, semantic consistency and diversity. Project page: https://boyuaner.github.io/ropar-main/

Bin Cao, Sipeng Zheng, Ye Wang et al.

Human motion generation has emerged as a critical technology with transformative potential for real-world applications. However, existing vision-language-motion models (VLMMs) face significant limitations that hinder their practical deployment. We identify controllability as a main bottleneck, manifesting in five key aspects: inadequate response to diverse human commands, limited pose initialization capabilities, poor performance on long-term sequences, insufficient handling of unseen scenarios, and lack of fine-grained control over individual body parts. To overcome these limitations, we present Being-M0.5, the first real-time, controllable VLMM that achieves state-of-the-art performance across multiple motion generation tasks. Our approach is built upon HuMo100M, the largest and most comprehensive human motion dataset to date, comprising over 5 million self-collected motion sequences, 100 million multi-task instructional instances, and detailed part-level annotations that address a critical gap in existing datasets. We introduce a novel part-aware residual quantization technique for motion tokenization that enables precise, granular control over individual body parts during generation. Extensive experimental validation demonstrates Being-M0.5's superior performance across diverse motion benchmarks, while comprehensive efficiency analysis confirms its real-time capabilities. Our contributions include design insights and detailed computational analysis to guide future development of practical motion generators. We believe that HuMo100M and Being-M0.5 represent significant advances that will accelerate the adoption of motion generation technologies in real-world applications. The project page is available at https://beingbeyond.github.io/Being-M0.5.

Yuting Mei, Ye Wang, Sipeng Zheng et al.

As robotic agents increasingly assist humans in reality, quadruped robots offer unique opportunities for interaction in complex scenarios due to their agile movement. However, building agents that can autonomously navigate, adapt, and respond to versatile goals remains a significant challenge. In this work, we introduce QuadrupedGPT designed to follow diverse commands with agility comparable to that of a pet. The primary challenges addressed include: i) effectively utilizing multimodal observations for informed decision-making; ii) achieving agile control by integrating locomotion and navigation; iii) developing advanced cognition to execute long-term objectives. Our QuadrupedGPT interprets human commands and environmental contexts using a large multimodal model. Leveraging its extensive knowledge base, the agent autonomously assigns parameters for adaptive locomotion policies and devises safe yet efficient paths toward its goals. Additionally, it employs high-level reasoning to decompose long-term goals into a sequence of executable subgoals. Through comprehensive experiments, our agent shows proficiency in handling diverse tasks and intricate instructions, representing a significant step toward the development of versatile quadruped agents for open-ended environments.

Sipeng Zheng, Bohan Zhou, Yicheng Feng et al.

In this paper, we propose \textbf{UniCode}, a novel approach within the domain of multimodal large language models (MLLMs) that learns a unified codebook to efficiently tokenize visual, text, and potentially other types of signals. This innovation addresses a critical limitation in existing MLLMs: their reliance on a text-only codebook, which restricts MLLM's ability to generate images and texts in a multimodal context. Towards this end, we propose a language-driven iterative training paradigm, coupled with an in-context pre-training task we term ``image decompression'', enabling our model to interpret compressed visual data and generate high-quality images.The unified codebook empowers our model to extend visual instruction tuning to non-linguistic generation tasks. Moreover, UniCode is adaptable to diverse stacked quantization approaches in order to compress visual signals into a more compact token representation. Despite using significantly fewer parameters and less data during training, Unicode demonstrates promising capabilities in visual reconstruction and generation. It also achieves performances comparable to leading MLLMs across a spectrum of VQA benchmarks.