视觉大模型 / VLM - arXivDaily 专题

2509.02659 2025-09-04 cs.CV cs.RO 83%

2nd Place Solution for CVPR2024 E2E Challenge: End-to-End Autonomous Driving Using Vision Language Model

Zilong Guo, Yi Luo, Long Sha, Dongxu Wang, Panqu Wang, Chenyang Xu, Yi Yang

机构 * ZERON Shanghai, China（上海零点科技有限公司）

专题命中 VLM训练与架构：vision language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments 2nd place in CVPR 2024 End-to-End Driving at Scale Challenge

2508.15903 2025-08-25 cs.CV 83%

VT-LVLM-AR: A Video-Temporal Large Vision-Language Model Adapter for Fine-Grained Action Recognition in Long-Term Videos

Kaining Li, Shuwei He, Zihan Xu

机构 * Xidian University（西电大学）

专题命中 VLM训练与架构：vision-language model(title,abstract)；LLaVA(abstract)；分类 cs.CV

2504.11695 2025-08-25 cs.CV cs.CL cs.MM 83%

Interpreting the linear structure of vision-language model embedding spaces

Isabel Papadimitriou, Huangyuan Su, Thomas Fel, Sham Kakade, Stephanie Gil

机构 * Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University（哈佛大学自然与人工智能研究所）； Department of Computer Science, Harvard University（哈佛大学计算机科学系）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments COLM 2025

2502.06860 2025-08-25 cs.CV cs.GR 83%

AutoSketch: VLM-assisted Style-Aware Vector Sketch Completion

Hsiao-Yuan Chin, I-Chao Shen, Yi-Ting Chiu, Ariel Shamir, Bing-Yu Chen

机构 * National Taiwan University（国立台湾大学）； The University of Tokyo（东京大学）； Reichman University（里奇曼大学）

专题命中 VLM训练与架构：VLM(title,abstract)；vision-language model(abstract)；分类 cs.CV

Comments 11 pages, Hsiao-Yuan Chin and I-Chao Shen contributed equally to the paper

2508.15688 2025-08-22 cs.CV 83%

LLM-empowered Dynamic Prompt Routing for Vision-Language Models Tuning under Long-Tailed Distributions

Yongju Jia, Jiarui Ma, Xiangxian Li, Baiqiao Zhang, Xianhui Cao, Juan Liu, Yulong Bian

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments accepted by EMNLP 2025

2508.13460 2025-08-20 cs.CV 83%

Revisiting MLLM Token Technology through the Lens of Classical Visual Coding

Jinming Liu, Junyan Lin, Yuntao Wei, Kele Shao, Keda Tao, Jianguo Huang, Xudong Yang, Zhibo Chen, Huan Wang, Xin Jin

机构 * Eastern Institute of Technology, Ningbo, China（东部技术研究所）； Westlake University（西湖大学）； USTC（中国科学技术大学）

专题命中 VLM训练与架构：MLLM(title,abstract)；multimodal large language model(abstract)；分类 cs.CV

2508.11317 2025-08-18 cs.CV cs.MM 83%

Logic Unseen: Revealing the Logical Blindspots of Vision-Language Models

Yuchen Zhou, Jiayu Tang, Shuo Yang, Xiaoyan Xiao, Yuqin Dai, Wenhao Yang, Chao Gou, Xiaobo Xia, Tat-Seng Chua

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

2508.10552 2025-08-15 cs.CL cs.AI 83%

When Language Overrules: Revealing Text Dominance in Multimodal Large Language Models

Huyu Wu, Meng Tang, Xinhan Zheng, Haiyun Jiang

专题命中 VLM训练与架构：multimodal large language model(title,abstract)；LLaVA(abstract)；分类 cs.AI

2504.04801 2025-08-15 cs.CV 83%

OrderChain: Towards General Instruct-Tuning for Stimulating the Ordinal Understanding Ability of MLLM

Jinhong Wang, Shuo Tong, Jian liu, Dongqi Tang, Weiqiang Wang, Wentong Li, Hongxia Xu, Danny Chen, Jintai Chen, Jian Wu

机构 * College of Computer Science & Technology, Zhejiang University（浙江大学计算机科学与技术学院）； Transvascular Implantation Devices Research Institute and Liangzhu Laboratory（血管植入物研究机构和良渚实验室）； Ant Group（蚂蚁集团）； University of Notre Dame（圣母大学）； HKUST (Guangzhou)（香港科技大学（广州））

专题命中 VLM训练与架构：MLLM(title)；LLaVA(abstract)；multimodal large language model(abstract)；分类 cs.CV

Comments Accepted by ICCV 2025

2505.17619 2025-08-14 cs.CV 83%

CAS-IQA: Teaching Vision-Language Models for Synthetic Angiography Quality Assessment

Bo Wang, De-Xing Huang, Xiao-Hu Zhou, Mei-Jiang Gui, Nu-Fang Xiao, Jian-Long Hao, Ming-Yuan Liu, Zeng-Guang Hou

机构 * State Key Laboratory of Multimodal Artificial Intelligence Systems（多模态人工智能系统国家重点实验室）； Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）； University of Chinese Academy of Sciences（中国科学院大学）； University of Science and Technology Beijing（北京科技大学）； Hunan University of Science and Technology（湖南科技大学）； Shanxi University of Finance and Economics（山西财经大学）； Beijing Friendship Hospital, Capital Medical University（首都医科大学北京友谊医院）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments Camera ready version for ICONIP 2025

2504.19860 2025-08-14 cs.CV 83%

CoherenDream: Boosting Holistic Text Coherence in 3D Generation via Multimodal Large Language Models Feedback

Chenhan Jiang, Yihan Zeng, Dit-Yan Yeung

专题命中 VLM训练与架构：multimodal large language model(title,abstract)；MLLM(abstract)；分类 cs.CV

2508.07835 2025-08-12 cs.CV 83%

Effortless Vision-Language Model Specialization in Histopathology without Annotation

Jingna Qiu, Nishanth Jain, Jonas Ammeling, Marc Aubreville, Katharina Breininger

机构 * Friedrich-Alexander-Universität Erlangen-Nürnberg（弗里德里希-亚历山大-厄林根-纽伦堡大学）； Ingolstadt University of Applied Sciences（因戈尔施塔特应用科学大学）； Flensburg University of Applied Sciences（弗拉森堡应用科学大学）； Julius-Maximilians-Universität Würzburg（朱利叶斯-马克斯-魏扎克大学）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

2502.11859 2025-08-06 cs.CV cs.CL 83%

Defining and Evaluating Visual Language Models' Basic Spatial Abilities: A Perspective from Psychometrics

Wenrui Xu, Dalin Lyu, Weihang Wang, Jie Feng, Chen Gao, Yong Li

机构 * School of Architecture, Tsinghua University（清华大学建筑学院）； Department of Electronic Engineering, Tsinghua University（清华大学电子工程系）； BNRist, Tsinghua University（清华大学BNRist）

专题命中 VLM训练与架构：visual language model(title,abstract)；VLM(abstract)；分类 cs.CV

Journal ref Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics. Volume 1: Long Papers (2025) 11571-11590

2508.02525 2025-08-05 cs.AI 83%

Accurate and Interpretable Postmenstrual Age Prediction via Multimodal Large Language Model

Qifan Chen, Jin Cui, Cindy Duan, Yushuo Han, Yifei Shi

机构 * King’s College London（伦敦国王学院）； Imperial College London（帝国理工学院）； Columbia University（哥伦比亚大学）

专题命中 VLM训练与架构：multimodal large language model(title,abstract)；MLLM(abstract)；分类 cs.AI

Comments Submitted to the NeurIPS 2025 Workshop GenAI4Health. Conference website: https://aihealth.ischool.utexas.edu/GenAI4HealthNeurips2025/

2508.02047 2025-08-05 cs.CV 83%

Mapillary Vistas Validation for Fine-Grained Traffic Signs: A Benchmark Revealing Vision-Language Model Limitations

Sparsh Garg, Abhishek Aich

机构 * NEC Laboratories, America（美国 NEC 实验室）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments Accepted to ICCV 2025 Workshop (4th DataCV Workshop and Challenge)

2508.01558 2025-08-05 cs.CV 83%

EvoVLMA: Evolutionary Vision-Language Model Adaptation

Kun Ding, Ying Wang, Shiming Xiang

机构 * Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments This paper has been accepted by ACM Multimedia 2025 (ACM MM 2025)

2507.22431 2025-07-31 cs.CV 83%

HQ-CLIP: Leveraging Large Vision-Language Models to Create High-Quality Image-Text Datasets and CLIP Models

Zhixiang Wei, Guangting Wang, Xiaoxiao Ma, Ke Mei, Huaian Chen, Yi Jin, Fengyun Rao

机构 * University of Science and Technology of China（中国科学技术大学）； WeChat Vision, Tencent Inc.（腾讯公司）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

2507.13568 2025-07-30 cs.CV 83%

LoRA-Loop: Closing the Synthetic Replay Cycle for Continual VLM Learning

Kaihong Wang, Donghyun Kim, Margrit Betke

机构 * Waymo ； Boston University（波士顿大学）； Korea University（韩国大学）

专题命中 VLM训练与架构：VLM(title,abstract)；vision-language model(abstract)；分类 cs.CV

2507.20519 2025-07-29 cs.CV 83%

AgroBench: Vision-Language Model Benchmark in Agriculture

Risa Shinoda, Nakamasa Inoue, Hirokatsu Kataoka, Masaki Onishi, Yoshitaka Ushiku

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments ICCV 2025

2507.20749 2025-07-29 cs.CL cs.CV 83%

Investigating Structural Pruning and Recovery Techniques for Compressing Multimodal Large Language Models: An Empirical Study

Yiran Huang, Lukas Thede, Massimiliano Mancini, Wenjia Xu, Zeynep Akata

机构 * Technical University of Munich, Germany（慕尼黑技术大学，德国）； Helmholtz Munich, Munich Center for Machine Learning, Germany（海德堡慕尼黑，慕尼黑机器学习中心，德国）； University of Tübingen, Tübingen AI Center, Germany（图宾根大学，图宾根人工智能中心，德国）； University of Trento, Italy（特伦托大学，意大利）； Beijing University of Posts and Telecommunications, China（北京邮电大学，中国）

专题命中 VLM训练与架构：multimodal large language model(title,abstract)；LLaVA(abstract)；分类 cs.CV

Comments Accepted at GCPR 2025

2507.15480 2025-07-29 cs.CV 83%

One Last Attention for Your Vision-Language Model

Liang Chen, Ghazi Shazan Ahmad, Tianjun Yao, Lingqiao Liu, Zhiqiang Shen

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments Accepted by ICCV 2025

2411.17762 2025-07-29 cs.CV 83%

MUSE-VL: Modeling Unified VLM through Semantic Discrete Encoding

Rongchang Xie, Chen Du, Ping Song, Chang Liu

专题命中 VLM训练与架构：VLM(title)；vision-language model(abstract)；LLaVA(abstract)；分类 cs.CV

Comments ICCV 2025

2507.11155 2025-07-16 cs.CR cs.AI 83%

Bridging the Gap in Vision Language Models in Identifying Unsafe Concepts Across Modalities

Yiting Qu, Michael Backes, Yang Zhang

机构 * CISPA Helmholtz Center for Information Security（CISPA海德堡信息安全研究中心）

专题命中 VLM训练与架构：vision language model(title)；vision-language model(abstract)；VLM(abstract)；分类 cs.AI

Comments To Appear in the 34th USENIX Security Symposium, August 2025

2412.18675 2025-07-16 cs.CV 83%

TAB: Transformer Attention Bottlenecks enable User Intervention and Debugging in Vision-Language Models

Pooyan Rahmanzadehgervi, Hung Huy Nguyen, Rosanne Liu, Long Mai, Anh Totti Nguyen

机构 * Auburn University（奥本大学）； Google DeepMind, ML Collective（谷歌DeepMind及ML集体）； Adobe Research（Adobe研究）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

2507.10355 2025-07-15 cs.CV 83%

Beyond Graph Model: Reliable VLM Fine-Tuning via Random Graph Adapter

Bo Jiang, Xueyang Ze, Beibei Wang, Xixi Wang, Xixi Wan, Bin Luo

机构 * Anhui Provincial Key Laboratory of Multimodal Cognitive Computation（安徽省级多模态认知计算重点实验室）； School of Computer Science and Technology（计算机科学与技术学院）； Anhui University（安徽大学）

专题命中 VLM训练与架构：VLM(title,abstract)；vision-language model(abstract)；分类 cs.CV

2507.09097 2025-07-15 cs.CV 83%

RadEyeVideo: Enhancing general-domain Large Vision Language Model for chest X-ray analysis with video representations of eye gaze

Yunsoo Kim, Jinge Wu, Honghan Wu

机构 * Institute of Health Informatics（健康信息学研究所）

专题命中 VLM训练与架构：vision language model(title)；vision-language model(abstract)；LLaVA(abstract)；分类 cs.CV

2503.09394 2025-07-15 cs.CV 83%

Bidirectional Prototype-Reward co-Evolution for Test-Time Adaptation of Vision-Language Models

Xiaozhen Qiao, Peng Huang, Jiakang Yuan, Xianda Guo, Bowen Ye, Chaocan Xue, Ye Zheng, Zhe Sun, Xuelong Li

机构 * School of Information Science and Technology, University of Science and Technology of China（信息科学与技术学院，中国科学技术大学）； Institute of Artificial Intelligence (TeleAI), China Telecom（人工智能研究所（TeleAI），中国电信）； College of Future Information Technology, Fudan University（未来信息技术学院，复旦大学）； College of Computer Science, Wuhan University（计算机科学学院，武汉大学）

专题命中 VLM训练与架构：vision-language model(title)；vision language model(abstract)；VLM(abstract)；分类 cs.CV

2503.20612 2025-07-08 cs.CV 83%

IAP: Improving Continual Learning of Vision-Language Models via Instance-Aware Prompting

Hao Fu, Hanbin Zhao, Jiahua Dong, Henghui Ding, Chao Zhang, Hui Qian

机构 * Zhejiang University（浙江大学）； Fudan University（复旦大学）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments Code can be found at https://github.com/FerdinandZJU/IAP

2507.02957 2025-07-08 cs.CV 83%

CS-VLM: Compressed Sensing Attention for Efficient Vision-Language Representation Learning

Andrew Kiruluta, Preethi Raju, Priscilla Burity

机构 * Berkeley（伯克利）

专题命中 VLM训练与架构：VLM(title)；vision-language model(abstract)；grounding(abstract)；分类 cs.CV

2507.00505 2025-07-08 cs.CV 83%

LLaVA-SP: Enhancing Visual Representation with Visual Spatial Tokens for MLLMs

Haoran Lou, Chunxiao Fan, Ziyan Liu, Yuexin Wu, Xinliang Wang

机构 * Beijing University of Posts and Telecommunications（北京邮电大学）； Beihang University（北航）

专题命中 VLM训练与架构：LLaVA(title,abstract)；multimodal large language model(abstract)；分类 cs.CV

Comments Accepted to ICCV 2025

1. VLM训练与架构 5030 篇

2nd Place Solution for CVPR2024 E2E Challenge: End-to-End Autonomous Driving Using Vision Language Model

VT-LVLM-AR: A Video-Temporal Large Vision-Language Model Adapter for Fine-Grained Action Recognition in Long-Term Videos

Interpreting the linear structure of vision-language model embedding spaces

AutoSketch: VLM-assisted Style-Aware Vector Sketch Completion

LLM-empowered Dynamic Prompt Routing for Vision-Language Models Tuning under Long-Tailed Distributions

Revisiting MLLM Token Technology through the Lens of Classical Visual Coding

Logic Unseen: Revealing the Logical Blindspots of Vision-Language Models

When Language Overrules: Revealing Text Dominance in Multimodal Large Language Models

OrderChain: Towards General Instruct-Tuning for Stimulating the Ordinal Understanding Ability of MLLM

CAS-IQA: Teaching Vision-Language Models for Synthetic Angiography Quality Assessment

CoherenDream: Boosting Holistic Text Coherence in 3D Generation via Multimodal Large Language Models Feedback

Effortless Vision-Language Model Specialization in Histopathology without Annotation

Defining and Evaluating Visual Language Models' Basic Spatial Abilities: A Perspective from Psychometrics

Accurate and Interpretable Postmenstrual Age Prediction via Multimodal Large Language Model

Mapillary Vistas Validation for Fine-Grained Traffic Signs: A Benchmark Revealing Vision-Language Model Limitations

EvoVLMA: Evolutionary Vision-Language Model Adaptation

HQ-CLIP: Leveraging Large Vision-Language Models to Create High-Quality Image-Text Datasets and CLIP Models

LoRA-Loop: Closing the Synthetic Replay Cycle for Continual VLM Learning

AgroBench: Vision-Language Model Benchmark in Agriculture

Investigating Structural Pruning and Recovery Techniques for Compressing Multimodal Large Language Models: An Empirical Study

One Last Attention for Your Vision-Language Model

MUSE-VL: Modeling Unified VLM through Semantic Discrete Encoding

Bridging the Gap in Vision Language Models in Identifying Unsafe Concepts Across Modalities

TAB: Transformer Attention Bottlenecks enable User Intervention and Debugging in Vision-Language Models

Beyond Graph Model: Reliable VLM Fine-Tuning via Random Graph Adapter

RadEyeVideo: Enhancing general-domain Large Vision Language Model for chest X-ray analysis with video representations of eye gaze

Bidirectional Prototype-Reward co-Evolution for Test-Time Adaptation of Vision-Language Models

IAP: Improving Continual Learning of Vision-Language Models via Instance-Aware Prompting

CS-VLM: Compressed Sensing Attention for Efficient Vision-Language Representation Learning

LLaVA-SP: Enhancing Visual Representation with Visual Spatial Tokens for MLLMs