Progressive Prompt Detailing for Improved Alignment in Text-to-Image Generative Models
机构 * Boston University（波士顿大学） ； Johns Hopkins University（约翰霍普金斯大学）
Comments Accepted at CVPR 2025 workshops (AI4CC (oral) & GMCV (poster))
期刊&会议
CVPR
Conference on Computer Vision and Pattern Recognition · 会议 · Computer Vision
HEIE: MLLM-Based Hierarchical Explainable AIGC Image Implausibility Evaluator
机构 * Tsinghua University（清华大学） ； BNRist ； OPPO AI Center（OPPO人工智能中心） ； Peking University（北京大学） ； Hangzhou Zhuoxi Institute of Brain and Intelligence（杭州智行脑科学与人工智能研究所）
Comments Accepted by CVPR 2025
Adventurer: Optimizing Vision Mamba Architecture Designs for Efficiency
机构 * Johns Hopkins University（约翰霍普金斯大学） ； UC Berkeley（伯克利大学） ； University of Florida（佛罗里达大学） ； UC Santa Cruz（圣克拉拉大学）
Comments Published in CVPR 2025
Mamba-R: Vision Mamba ALSO Needs Registers
机构 * Johns Hopkins University（约翰霍普金斯大学） ； University of Florida（佛罗里达大学） ； UC Santa Cruz（加州圣克拉拉大学）
Comments Published in CVPR 2025
MaskAdapt: Unsupervised Geometry-Aware Domain Adaptation Using Multimodal Contextual Learning and RGB-Depth Masking
机构 * University of Regina（里贾纳大学） ； University Canada West（加拿大西部大学） ； The University of Western Australia（西澳大学）
Comments 11 pages, 5 figures, presented at the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2025. Reviewer comments available upon request
Argus: Vision-Centric Reasoning with Grounded Chain-of-Thought
机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； NVIDIA
Comments CVPR 2025. Project Page: https://yunzeman.github.io/argus/
Sketch Down the FLOPs: Towards Efficient Networks for Human Sketch
机构 * SketchX, CVSSP, University of Surrey, United Kingdom（SketchX、CVSSP、塞夫顿大学、英国） ； Department of Computer Science, University of Central Florida（计算机科学系、中央佛罗里达大学）
Comments Accepted at CVPR 2025, Project Page: https://subhajitmaity.me/SketchDownTheFLOPs
Boosting Domain Incremental Learning: Selecting the Optimal Parameters is All You Need
机构 * Xi’an Jiaotong University（西安交通大学） ； Shenzhen University of Advanced Technology（深圳先进技术大学）
Comments Accepted at CVPR 2025
Bridging Classical and Modern Computer Vision: PerceptiveNet for Tree Crown Semantic Segmentation
机构 * University of Oxford（牛津大学）
Comments Accepted for publication at the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) EarthVision
Synthetic Generation and Latent Projection Denoising of Rim Lesions in Multiple Sclerosis
Comments Accepted full paper in Synthetic Data @ CVPR 2025 12 pages, 10 figures
Wav2Sem: Plug-and-Play Audio Semantic Decoupling for 3D Speech-Driven Facial Animation
机构 * Beihang University（北京航空航天大学） ； Peng Cheng Laboratory（鹏城实验室） ； North China University of Technology（华北理工大学） ； University of Washington（华盛顿大学） ； University of Copenhagen（哥本哈根大学）
Comments Accepted to CVPR 2025
Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction Networks for Single-Pixel Imaging
机构 * Westlake University（西湖大学） ； Zhejiang University（浙江大学） ； Shanghai Jiao Tong University（上海交通大学）
Comments Accepted by CVPR 2025
URWKV: Unified RWKV Model with Multi-state Perspective for Low-light Image Restoration
机构 * Fujian Key Laboratory of Network Computing and Intelligent Information Processing, College of Computer and Data Science, Fuzhou University（福建网络计算与智能信息处理重点实验室，计算机与数据科学学院，福州大学） ； Engineering Research Center of Big Data Intelligence, Ministry of Education（大数据智能工程研究中心，教育部）
Comments This paper has been accepted to CVPR 2025
4DTAM: Non-Rigid Tracking and Mapping via Dynamic Surface Gaussians
机构 * Dyson Robotics Laboratory, Imperial College London（帝京理工学院伦敦分校动态机器人实验室）
Comments CVPR 2025. Project Page: https://muskie82.github.io/4dtam/
A Probabilistic Jump-Diffusion Framework for Open-World Egocentric Activity Recognition
机构 * CSSE Department, Auburn University（安全科学与工程系，阿伯丁大学）
Comments Extended abstract of arXiv:2504.03948 for CVPR 2025 EgoVis Workshop
It's a (Blind) Match! Towards Vision-Language Correspondence without Parallel Data
机构 * TU Munich（慕尼黑技术大学） ； Munich Center for Machine Learning（慕尼黑机器学习中心）
Comments Accepted to CVPR 2025, Project page: https://dominik-schnaus.github.io/itsamatch/
From Head to Tail: Towards Balanced Representation in Large Vision-Language Models through Adaptive Data Calibration
机构 * Fudan University（复旦大学） ； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室） ； Stony Brook University（石溪大学） ； The Chinese University of Hong Kong（香港中文大学）
Comments Accepted by CVPR 2025. Project Page: https://vlmlt.github.io/
SeeGround: See and Ground for Zero-Shot Open-Vocabulary 3D Visual Grounding
机构 * HKUST(GZ)（香港科技大学（广州）） ； I 2 R, A*STAR（I2R, A*STAR） ； National University of Singapore（新加坡国立大学） ； CSE, HKUST（香港科技大学计算机科学与工程系）
Comments CVPR 2025; 21 pages, 10 figures, 10 tables; Code at https://seeground.github.io/
Circumventing shortcuts in audio-visual deepfake detection datasets with unsupervised learning
机构 * Bitdefender ； Politehnica Bucharest（巴尔干理工大学）
Comments Accepted as a highlight paper at the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2025
SynTable: A Synthetic Data Generation Pipeline for Unseen Object Amodal Instance Segmentation of Cluttered Tabletop Scenes
机构 * Advanced Robotics Centre, National University of Singapore（新加坡国立大学先进机器人中心）
Comments Camera-ready version for SynData4CV Workshop @ CVPR 2025. 18 Pages, 11 figures
PS4PRO: Pixel-to-pixel Supervision for Photorealistic Rendering and Optimization
机构 * School of Electrical and Computer Engineering, Purdue University（电子与计算机工程学院，普渡大学）
Comments Accepted to the CVPR 2025 Workshop on Autonomous Driving (WAD)
Zero-Shot 3D Visual Grounding from Vision-Language Models
机构 * HKUST(GZ)（香港科技大学（广州）） ； I 2 R, A*STAR（I2R, A*STAR） ； NUS（国立大学） ； CSE, HKUST（计算机科学与工程系，香港科技大学）
Comments 3D-LLM/VLA @ CVPR 2025; Project Page at https://seeground.github.io/
RC-AutoCalib: An End-to-End Radar-Camera Automatic Calibration Network
机构 * National Yang Ming Chiao Tung University（国立阳明交通大学） ； Ho Chi Minh City University of Technology and Education（胡志明市技术与教育大学）
Comments Accepted to CVPR 2025
Bringing CLIP to the Clinic: Dynamic Soft Labels and Negation-Aware Learning for Medical Analysis
机构 * Interdisciplinary Program in Bioengineering, Seoul National University Graduate School（生物工程跨学科项目，首尔国立大学研究生院） ； Integrated Major in Innovative Medical Science, Seoul National University Graduate School（创新医学联合专业，首尔国立大学研究生院） ； Department of Radiology, Seoul National University Hospital（放射科，首尔国立大学医院）
Comments 16 pages (8 main, 2 references, 6 appendix), 13 figures. Accepted to CVPR 2025. This author-accepted manuscript includes an expanded ethics/data user agreement section. The final version will appear in the Proceedings of CVPR 2025
Point-to-Region Loss for Semi-Supervised Point-Based Crowd Counting
机构 * Department of Computer Science, City University of Hong Kong（计算机科学系，香港城市大学）
Comments accepted by CVPR-2025(highlight)
Knowledge Distillation Approach for SOS Fusion Staging: Towards Fully Automated Skeletal Maturity Assessment
机构 * Department of Electrical and Computer Engineering, University of Illinois Chicago（伊利诺伊大学芝加哥分校电气与计算机工程系） ； Department of Orthodontics, College of Dentistry, University of Illinois Chicago（伊利诺伊大学芝加哥分校牙科学院正畸学系） ； University of Orleans（奥尔良大学）
Comments This paper has been accepted to the CVPR Workshop 2025, to be held in Nashville, Tennessee
Enhancing Vision Transformer Explainability Using Artificial Astrocytes
机构 * University of Miami（迈阿密大学） ； University of A Coruña（阿罗乌纳大学）
Comments LXCV Workshop at IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) 2025
Dual-Path Enhancements in Event-Based Eye Tracking: Augmented Robustness and Adaptive Temporal Modeling
机构 * University of Science, VNU-HCM（越南胡志明市国家大学） ； Vietnam National University（越南国家大学）
Comments Camera-ready version for CVPRW 2025. Accepted for presentation at the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW 2025)
Journal ref IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Nashville, 2025
Functionality understanding and segmentation in 3D scenes
机构 * Fondazione Bruno Kessler（布鲁诺·凯塞勒基金会） ； University of Trento（特伦托大学）
Comments CVPR 2025 Highlight. Camera ready version. 20 pages, 12 figures, 7 tables. Fixed typo in Eq.2
Empowering Vector Graphics with Consistently Arbitrary Viewing and View-dependent Visibility
机构 * School of Artificial Intelligence, University of Chinese Academy of Sciences（人工智能学院，中国科学院大学）
Comments CVPR 2025
