Researcher — paper.dinq.me

Segment Anything Model 3 (SAM3) advances open-vocabulary segmentation through promptable concept segmentation, enabling users to segment all instances associated with a given concept using short noun-phrase (NP) prompts. While effective for concept-level grounding, real-world interactions often i...

可解释性多模态推理OpenReview 项目页

作者位次 #8/10ACL 2026Findings· Findings of ACL

Seek-and-Solve: Benchmarking MLLM s for Visual Clue-Driven Reasoning in Daily Scenarios

Xiaomin Li · Tala Wang · Zichen Zhong · Ying Zhang · Zirui Zheng · Takashi Isobe · Dezhuang Li · Huchuan Lu · You He · XU JIA

摘要

Daily scenarios are characterized by visual richness, requiring Multimodal Large Language Models (MLLMs) to filter noise and identify decisive visual clues for accurate reasoning. Yet, current benchmarks predominantly aim at evaluating MLLMs’ pre-existing knowledge or perceptual understanding, of...

多模态评测与基准推理OpenReview 项目页

作者位次 #10/11SIGGRAPH 2026Oral· Computer Graphics (Conference Track)

VFXMaster: Unlocking Dynamic Visual Effect Generation via In-Context Learning

Baolu Li · Yiming Zhang · Qinghe Wang · Liqian Ma · Xiaoyu Shi · Xintao WANG · Pengfei Wan · Zhenfei Yin · Yunzhi Zhuge · Huchuan Lu · XU JIA

多模态理论与学习动力学OpenReview

作者位次 #4/4ICRA 2026Oral· Robotics & Automation

Eva-Tracker: ESDF-update-free, Visibility-aware Planning with Target Reacquisition for Robust Aerial Tracking

Yue Lin · Yang Liu · Dong Wang · Huchuan Lu

推理OpenReview

作者位次 #6/6IROS 2026Poster· Ensuring Safety in Robot Motion and Control

UCON: Uncertainty-Aware Navigation with Historical Re-Association in Dynamic Environments

Bing Sun · Yue Lin · Yongsheng Yuan · Yang Liu · Dong Wang · Huchuan Lu

推理安全Collision AvoidanceMotion and Path PlanningReactive and Sensor-Based PlanningOpenReview

作者位次 #7/7AAAI 2026Poster· Artificial Intelligence

CADTrack: Learning Contextual Aggregation with Deformable Alignment for Robust RGBT Tracking

Hao Li · Yuhao Wang · Xiantao Hu · Wenning Hao · Pingping Zhang · Dong Wang · Huchuan Lu

对齐与偏好OpenReview

作者位次 #4/4AAAI 2026Poster· Artificial Intelligence

X-ReID: Multi-granularity Information Interaction for Video-Based Visible-Infrared Person Re-Identification

Chenyang Yu · Xuehu Liu · Pingping Zhang · Huchuan Lu

多模态OpenReview

Huchuan Lu

获取完整研究档案

联系方式

2026 顶会论文

RELO: Reinforcement Learning to Localize for Visual Object Tracking

StructMamPose: From Sequential Perception to Structural Reasoning for 3D Human Pose Estimation

RAGTrack: Language-aware RGBT Tracking with Retrieval-Augmented Generation

TGTrack: Temporal Generative Learning for Unified Single Object Tracking

Reinforcing Video Object Segmentation to Think before it Segments

MultiShotMaster: A Controllable Multi-Shot Video Generation Framework

Ego-InBetween: Generating Object State Transitions in Ego-Centric Videos

UETrack: A Unified and Efficient Framework for Single Object Tracking

Complementary Prototype Mapping for Efficient Multimodal Anomaly Detection

UniMMAD: Unified Multi-Modal and Multi-Class Anomaly Detection via MoE-Driven Feature Decompression

SAM 3- I : Segment Anything with Instructions