Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Suggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works96 from public data
- Efficient and Accurate Arbitrary-Shaped Text Detection With Pixel Aggregation Network632
This paper proposes an efficient and accurate arbitrary-shaped text detector, termed Pixel Aggregation Network (PAN), which is equipped with a low computational-cost segmentation head and a learnable post-processing.
- VLMEvalKit: An Open-Source ToolKit for Evaluating Large Multi-Modality Models628
The VLMEvalKit toolkit aims to provide a user-friendly and comprehensive framework for researchers and developers to evaluate existing multi-modality models and publish reproducible evaluation results.
- Visual-RFT: Visual Reinforcement Fine-Tuning575
This work introduces Visual Reinforcement Fine-Tuning (Visual-RFT), which further extends the application areas of RFT on visual tasks and represents a paradigm shift in fine-tuning LVLMs, offering a data-efficient, rewarddriven approach that enhances reasoning and adaptability for domain-specific tasks.
- InternLM2 Technical Report440
InternLM2 is introduced, an open-source LLM that outperforms its predecessors in comprehensive evaluations across 6 dimensions and 30 benchmarks, long-context modeling, and open-ended subjective evaluations through innovative pre-training and optimization techniques.
- InternLM-XComposer2: Mastering Free-form Text-Image Composition and Comprehension in Vision-Language Large Model397
Experimental results demonstrate the superiority of InternLM-XComposer2 based on InternLM2-7B in producing high-quality long-text multi-modal content and its exceptional vision-language understanding performance across various benchmarks, where it not only significantly outperforms existing multimodal models but also matches or even surpasses GPT-4V and Gemini Pro in certain assessments.
- Open-Vocabulary DETR with Conditional Matching336
This work proposes a novel open-vocabulary detector based on DETR -- hence the name OV-DETR, which, once trained, can detect any object given its class name or an exemplar image and achieves non-trivial improvements over current state of the arts.
- PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction332
This work proposes PyramidDrop, a visual redundancy reduction strategy for LVLMs to boost their efficiency in both training and inference with neglectable performance loss, and could also serve as a plug-and-play strategy for inference acceleration without training, with better performance and lower inference cost than counterparts.
- Seesaw Loss for Long-Tailed Instance Segmentation314
This work proposes Seesaw Loss to dynamically re-balance gradients of positive and negative samples for each category, with two complementary factors, i.e., mitigation factor and compensation factor, which obtains significant gains over Cross-Entropy Loss and achieves state-of-the-art performance on LVIS dataset without bells and whistles.
- Scene Text Detection with Supervised Pyramid Context Network238
A supervised pyramid context network (SPCNET) is proposed to precisely locate text regions while suppressing false positives in scene text detection, based on Feature Pyramid Network (FPN) and instance segmentation.
- Alpha-CLIP: A CLIP Model Focusing on Wherever you Want223
Alpha-CLIp, an enhanced version of CLIP with an auxiliary alpha channel to suggest attentive regions and fine-tuned with constructed millions of RGBA region-text pairs, is introduced, which not only preserves the visual recognition ability of CLIP but also enables precise control over the emphasis of image contents.
- InternLM-XComposer-2.5: A Versatile Large Vision Language Model Supporting Long-Contextual Input and Output216
A versatile large-vision language model that supports long-contextual input and output, InternLM-XComposer-2.5 excels in various text-image comprehension and composition applications, achieving GPT-4V level capabilities with merely 7B LLM backend.
- Unified Vision and Language Prompt Learning213
A systematic study on two representative prompt tuning methods, namely text prompt tuning and visual prompt tuning, and proposes a simple approach called Unified Prompt Tuning (UPT), which essentially learns a tiny neural network to jointly optimize prompts across different modalities.
- Contextual Object Detection with Multimodal Large Language Models182
This work presents ContextDET, a unified multimodal model that is capable of end-to-end differentiable modeling of visual-language contexts, so as to locate, identify, and associate visual objects with language inputs for human-AI interaction.
- OVO-Bench: How Far is Your Video-LLMs from Real-World Online Video Understanding?159
OVO-Bench evaluates the ability of video LLMs to reason and respond to events occurring at specific timestamps under three distinct scenarios: Backward tracing, real-time understanding, and forward active responding.
- FASA: Feature Augmentation and Sampling Adaptation for Long-Tailed Instance Segmentation154
This work proposes a simple yet effective method, Feature Augmentation and Sampling Adaptation (FASA), that addresses the data scarcity issue by augmenting the feature space especially for rare classes by eliminating the need for inter-class transfer learning.
- Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction146
Experiments prove that Dispider outperforms existing methods not only in its superior understanding of video content in conventional video QA settings, but also in proactive response capability and temporal awareness under the streaming setting.
- 134
- MVSGaussian: Fast Generalizable Gaussian Splatting Reconstruction from Multi-View Stereo125
A new generalizable 3D Gaussian representation approach derived from Multi-View Stereo that can efficiently reconstruct unseen scenes with convincing generalizability, real-time rendering speed, and fast per-scene optimization.
- MotionClone: Training-Free Motion Cloning for Controllable Video Generation124
MotionClone is proposed, a training-free framework that enables motion cloning from reference videos to versatile motion-controlled video generation, including text-to-video and image-to-video, with notable superiority in terms of motion fidelity, textual alignment, and temporal consistency.
- 111
- Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning105
This work proposes Pref-GRPO, a pairwise preference reward-based GRPO method that shifts the optimization objective from score maximization to preference fitting, ensuring more stable training and mitigating reward hacking.
- 103
- Visionary-R1: Mitigating Shortcuts in Visual Reasoning with Reinforcement Learning94
This paper aims to train visual language models to perform reasoning on image data through reinforcement learning and visual question-answer pairs, without any explicit chain-of-thought (CoT) supervision, and trains the model to adhere to a caption-reason-answer output format.
- VideoRoPE: What Makes for Good Video Rotary Position Embedding?71
VideoRoPE is introduced, with a 3D structure designed to preserve spatio-temporal relationships, which consistently surpasses previous RoPE variants, across diverse downstream tasks such as long video retrieval, video understanding, and video hallucination.
- 65
- SIM-CoT: Supervised Implicit Chain-of-Thought61
SIM-CoT significantly improves both in-domain accuracy and out-of-domain stability of implicit CoT methods, boosting Coconut by +8.2\% on GPT-2 and CODI by +3.0\% on LLaMA-3.1 8B.
- SongGen: A Single Stage Auto-regressive Transformer for Text-to-Song Generation57
The proposed SongGen model facilitates fine-grained control over diverse musical attributes, including lyrics and textual descriptions of instrumentation, genre, mood, and timbre, while also offering an optional three-second reference clip for voice cloning.
- InternLM-XComposer2.5-OmniLive: A Comprehensive Multimodal System for Long-term Streaming Video and Audio Interactions52
This project draws inspiration from the concept of the Specialized Generalist AI and introduces disentangled streaming perception, reasoning, and memory mechanisms, enabling real-time interaction with streaming video and audio input.
- MM-IFEngine: Towards Multimodal Instruction Following49
The MMIFEngine pipeline is presented, an effective pipeline to generate high-quality image-instruction pairs and a comprehensive evaluation pipeline incorporating both rule-based assessment and judge model, and introduces MM-IFEval, a challenging and diverse multi-modal instruction-following benchmark.
- Visual Agentic Reinforcement Fine-Tuning46
The findings suggest that Visual-ARFT offers a promising path toward building robust and generalizable multimodal agents.
- Spatial-SSRL: Enhancing Spatial Understanding via Self-Supervised Reinforcement Learning44
Spatial-SSRL is introduced, a self-supervised RL paradigm that derives verifiable signals directly from ordinary RGB or RGB-D images and provides a practical route to stronger spatial intelligence in LVLMs.
- Light-a-Video: Training-Free Video Relighting via Progressive Light Fusion44
Experiments show that Light-AVideo improves the temporal consistency of relighted video while maintaining the relighted image quality, ensuring coherent lighting transitions across frames.
- Beyond Fixed: Training-Free Variable-Length Denoising for Diffusion Large Language Models36
DAEDAL is introduced, a novel training-free denoising strategy that enables Dynamic Adaptive Length Expansion for Diffusion Large Language Models and achieves performance comparable, and in some cases superior, to meticulously tuned fixed-length baselines, while simultaneously enhancing computational efficiency by achieving a higher effective token ratio.
- 35
- WildClawBench: A Benchmark for Real-World, Long-Horizon Agent Evaluation34
WildClawBench, a native-runtime benchmark of 60 human-authored, bilingual, multimodal tasks spanning six thematic categories, shows that long-horizon, native-runtime agent evaluation remains a far-from-resolved task for current frontier models.
- MIA-DPO: Multi-Image Augmented Direct Preference Optimization For Large Vision-Language Models33
Multi-Image Augmented Direct Preference Optimization (MIA-DPO) mitigates the scarcity of diverse multi-image training data by extending single-image data with unrelated images arranged in grid collages or pic-in-pic formats, significantly reducing the costs associated with multi-image data annotations.
- RAR: Retrieving and Ranking Augmented MLLMs for Visual Recognition31
RAR, a Retrieving And Ranking augmented method for Multimodal Large Language Models, demonstrates a significant improvement in performance on 5 fine-grained visual recognition benchmarks, 11 few-shot image recognition datasets, and the 2 object detection datasets under the zero-shot recognition setting.
- 30
- 28
- UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generation25
UniGenBench++, a unified semantic assessment benchmark for T2I generation that comprehensively probes T2I models'santic consistency over 10 primary and 27 sub evaluation criteria, is introduced and an effective pipeline is developed for reliable benchmark construction and streamlined model assessment.
- 24
- Intern-S1-Pro: Scientific Multimodal Foundation Model at Trillion Scale23
This work introduces Intern-S1-Pro, the first one-trillion-parameter scientific multimodal foundation model, which further fortifies the fusion of general and specialized intelligence, working as a Specializable Generalist, demonstrating its position in the top tier of open-source models for general capabilities.
- BoostStep: Boosting mathematical capability of Large Language Models via improved single-step reasoning22
BoostStep is a method that enhances reasoning accuracy through step-aligned ICL, a novel mechanism that carefully aligns retrieved reference steps with the corresponding reasoning steps, and enhances state-of-the-art LLMs to solve challenging math problems using simpler examples.
- Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate22
This paper presents the Modality Integration Rate (MIR), an effective, robust, and generalized metric to indicate the multi-modal pre-training quality of Large Vision Language Models (LVLMs), and proposes evaluating the pre-training quality from the inter-modal distribution distance perspective.
- Overcoming the Pitfalls of Vision-Language Model Finetuning for OOD Generalization22
This paper demonstrates that vision-language models, after long enough finetuning but without proper regularization, tend to overfit the known classes in the given dataset, with degraded performance on unknown classes, and proposes a novel approach OGEN to address this pitfall.
- Unified Personalized Reward Model for Vision Generation17
This work proposes UnifiedReward-Flex, a unified personalized reward model for vision generation that couples reward modeling with flexible and context-adaptive reasoning behaviors, and integrates it into the GRPO framework for image and video synthesis, demonstrating its superiority.
- Bootstrapping Grounded Chain-of-Thought in Multimodal Llms for Data-Efficient Model Adaptation16
GCoT is proposed, a simple bootstrapping-based approach that aims to inject grounding information (i.e., bounding boxes) into CoT data, essentially making the reasoning steps more faithful to input images, and demonstrates that under data-limited regimes the approach significantly improves upon fine-tuning and distillation.
- 16
- Semi-Supervised and Long-Tailed Object Detection with CascadeMatch16
This paper proposes a novel pseudo-labeling-based detector called CascadeMatch, which surpasses existing state-of-the-art semi-supervised approaches—across a wide range of detection architectures—in handling long-tailed object detection.
- X-Prompt: Towards Universal In-Context Image Generation in Auto-Regressive Vision Language Foundation Models15
This work introduces X-Prompt, a purely auto-regressive large-vision language model designed to deliver competitive performance across a wide range of both seen and unseen image generation tasks, all within a unified in-context learning framework.
- ScaleCap: Inference-Time Scalable Image Captioning via Dual-Modality Debiasing14
ScaleCap is presented, an inference-time scalable image captioning strategy that generates comprehensive and detailed image captions that continuously enriches and calibrates the caption with increased inference budget and employs sentence-level offline contrastive decoding to effectively identify and eliminate hallucinations caused by linguistic biases.
- 12
- ARM-Thinker: Reinforcing Multimodal Generative Reward Models with Agentic Tool Use and Visual Reasoning11
ARM-Thinker is presented, an Agentic multimodal Reward Model that autonomously invokes external tools to ground judgments in verifiable evidence, replacing static, non-interactive reward scoring.
- 11
- Demo-ICL: In-Context Learning for Procedural Video Knowledge Acquisition10
Demo-ICL, an MLLM with a two-stage training strategy: video-supervised fine-tuning and information-assisted direct preference optimization, jointly enhancing the model's ability to learn from in-context examples is developed.
- SPARK: Synergistic Policy And Reward Co-Evolving Framework10
The Synergistic Policy And Reward Co-Evolving Framework (SPARK), an efficient, on-policy, and stable method that builds on RLVR, is introduced, an efficient, on-policy, and stable method that eliminates the need for a separate reward model and costly human preference data.
- 9
- CapRL++: Unified Reinforcement Learning with Verifiable Rewards for Dense Image and Video Captioning8
Results validate that CapRL++ effectively trains models to produce generalizable, high-fidelity descriptions, establishing a robust foundation beyond the limitations of traditional SFT.
- 8
- Trust Your Critic: Robust Reward Modeling and Reinforcement Learning for Faithful Image Editing and Generation8
Firm-reward.io presents FIRM (Faithful Image Reward Modeling), a comprehensive framework that develops robust reward models to provide accurate and reliable guidance for faithful image generation and editing and achieves substantial performance breakthroughs.
- EMemBench: Interactive Benchmarking of Episodic Memory for VLM Agents8
This work introduces EMemBench, a programmatic benchmark generator for evaluating long-term episodic memory of agents through interactive games, and evaluates memory agents with strong LMs/VLMs as backbones, using in-context prompting as baselines.
- On-Device Domain Generalization8
A systematic study of domain generalization (DG) for tiny neural networks finds that knowledge distillation (KD), a well-known technique for model compression, is much better for tackling the on-device DG problem than conventional DG methods.
- ByTheWay: Boost Your Text-to-Video Generation Model to Higher Quality in a Training-free Way7
ByTheWay is a training-free method to improve the quality of text-to-video generation without introducing additional parameters, augmenting memory or sampling time, and demonstrates that ByTheWay significantly improves the quality of text-to-video generation with negligible additional cost.
- 7
- KPNet: Towards Minimal Face Detector7
This work finds that the appearance feature of a generic face is discriminative enough for a tiny and shallow neural network to verify from the background, and achieves state-of-the-art accuracy on generic face detection and alignment benchmarks with only $\sim1M$ parameters.
- EndoCoT: Scaling Endogenous Chain-of-Thought Reasoning in Diffusion Models6
Endogenous Chain-of-Thought (EndoCoT) is proposed, a novel framework that first activates MLLMs'reasoning potential by iteratively refining latent thought states through an iterative thought guidance module, and then bridges these states to the DiT's denoising process.
- OVO-S-Bench: A Hierarchical Benchmark for Streaming Spatial Intelligence in Multimodal LLMs6
OVO-S-Bench is introduced, a fully human-annotated benchmark for streaming spatial intelligence, comprising 1,680 questions over 348 source videos, and it is found that chain-of-thought reasoning amplifies spatial errors when ungrounded in the stream.
- Visual Self-Refine: A Pixel-Guided Paradigm for Accurate Chart Parsing6
This work proposes a new paradigm named Visual Self-Refine (VSR), to enable a model to generate pixel-level localization outputs, visualize them, and then feed these visualizations back to itself, allowing it to intuitively inspect and correct its own potential visual perception errors.
- Bootstrap3D: Improving Multi-View Diffusion Model with Synthetic Data6
Bootstrap3D, a novel framework that automatically generates filtered multi-view images to assist in training multi-view diffusion models, is proposed and a Training Timestep Reschedule (TTR) strategy that leverages the denoising process to learn multi-view consistency while maintaining the original 2D diffusion prior is presented.
- TRivia: Self-supervised Fine-tuning of Vision-Language Models for Table Recognition6
TRivia is introduced, a self-supervised fine-tuning method that enables pretrained VLMs to learn TR directly from unlabeled table images in the wild and presents TRivia-3B, an open-sourced, compact, and state-of-the-art TR model that surpasses existing systems on three popular benchmarks.
- Unified Scene Representation and Reconstruction for 3D Large Language Models6
The importance of having a unified scene representation and reconstruction framework, which is essential for LLMs in 3D scenes, is demonstrated by introducing Uni3DR^2, which extracts 3D geometric and semantic aware representation features via the frozen pre-trained 2D foundation models and a multi-scale aggregate 3D decoder.
- 5
- Skill-as-Pseudocode: Refactoring Skill Libraries to Pseudocode for LLM Agents4
Skill-as-Pseudocode (SaP) is proposed, an automatic conversion of markdown skill libraries into typed pseudocode with deterministic quality control that wins paired games and falls below the prose baseline on the ALFWorld unseen split.
- WildAvatar: Learning In-the-wild 3D Avatars from the Web4
This paper curates WildAvatar, a web-scale in-the-wild human avatar creation dataset extracted from YouTube, with 10,000+ different human subjects and scenes, and demonstrates the quality and generalizability of avatar creation methods on WildAvatar.
- GraphThinker: Reinforcing Temporally Grounded Video Reasoning with Event Graph Thinking3
GraphThinker is proposed, a reinforcement finetuning-based method that constructs structural event-level scene graphs and enhances visual grounding to jointly reduce hallucinations in video reasoning.
- Generative Photographic Control for Scene-Consistent Video Cinematic Editing3
This paper proposes CineCtrl, the first video cinematic editing framework that provides fine control over professional camera parameters (e.g., bokeh, shutter speed), and introduces a decoupled cross-attention mechanism to disentangle camera motion from photographic inputs, allowing fine-grained, independent control without compromising scene consistency.
- SetCon: Towards Open-Ended Referring Segmentation via Set-Level Concept Prediction2
This work reformulates open-ended referring segmentation as explicit set-level concept prediction and proposes Set-Concept Segmentation (SetCon), which uses LVLM-generated natural-language concepts, instead of segmentation-specific tokens, as semantic conditions for joint mask-set decoding.
- Pave-GRPO: Beyond Instantaneous Guidance through Principled Average Velocity Decomposition2
Pave-GRPO significantly broadens the effective optimization scope under a fixed sampling and reward budget and distributes reward signals across more intermediate stages of the denoising process, enabling finer-grained and more thorough preference optimization.
- AdaGRPO: A Capability-Aware Adaptive Enhancement for Flow-based GRPO2
Adaptive GRPO (AdaGRPO) is proposed, a novel capability-aware RL algorithm tailored for flow models that consistently drives performance gains while significantly stabilizes GRPO training for flow models.
- ETCHR: Editing To Clarify and Harness Reasoning2
This work introduces ETCHR (Editing To Clarify and Harness Reasoning), a question-conditioned, reasoning-aware image editor decoupled from the downstream understanding model and trained with a two-stage recipe targeted at the two gaps: Reasoning Imitation via supervised fine-tuning on edit trajectories, followed by Reasoning Enhancement with VLM-derived rewards for edit correctness and downstream reasoning accuracy.
- Visual-ERM: Reward Modeling for Visual Equivalence2
This work proposes Visual Equivalence Reward Model (Visual-ERM), a multimodal generative reward model that provides fine-grained, interpretable, and task-agnostic feedback to evaluate vision-to-code quality directly in the rendered visual space and suggests that fine-grained visual reward supervision is both necessary and sufficient for vision-to-code RL, regardless of task specificity.
- 1
- V3Det Challenge 2024 on Vast Vocabulary and Open Vocabulary Object Detection: Methods and Results1
The V3Det Challenge 2024 aims to push the boundaries of object detection research and encourage innovation in this field, and aims to inspire future research directions in vast vocabulary and open-vocabulary object detection, driving progress in this field.
- –
- –
- –
- –
- –
- –
- –
- Beyond the Current Observation: Evaluating Multimodal Large Language Models in Controllable Non-Markov Games–
This work introduces RNG-Bench (Reconstructive Non-Markov Games), a benchmark suite designed to isolate a base model's ability to reconstruct past observations and act on them during multi-step interaction, and fine-tuning Qwen3.5-9B on optimal-policy rollouts and filtered model demonstrations improves performance on RNG-Bench and transfers to existing benchmarks without degrading general multimodal capability.
- –
- –
- –
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.