Accepted Papers

• How Important are Videos for Training Video LLMs? (Oral)
  George Lydakis¹, Alexander Hermans¹, Ali Athar², Daan de Geus^1,3, Bastian Leibe^{1
  1}RWTH Aachen University
  ²ByteDance Seed
  ³Eindhoven University of Technology
• Zero-shot Action Localization via the Confidence of Large Vision-Language Models
  Josiah Aklilu¹, Xiaohan Wang¹, Serena Yeung-Levy^{1
  1}Stanford University
• FiLA-Video: Spatio-Temporal Compression for Fine-Grained Long Video Understanding
  Yanan Guo¹, Wenhui Dong², Jun Song², Shiding Zhu³, Xuan Zhang², Hanqing Yang², Yingbo Wang², Yang Du², Xianing Chen², Bo Zheng^{2
  1}University of Science and Technology of China
  ²Alibaba Group
  ³Zhejiang University
• RoadSocial: A Diverse VideoQA Dataset and Benchmark for Road Event Understanding from Social Video Narratives
  Chirag Parikh¹, Deepti Rawat¹, Rakshitha R. T.¹, Tathagata Ghosh¹, Ravi Kiran Sarvadevabhatla^{1
  1}CVIT & iHub-Data, IIIT Hyderabad, India
• VELOCITI: Benchmarking Video-Language Compositional Reasoning with Strict Entailment
  Darshana Saravanan¹, Varun Gupta¹, Darshan Singh¹, Zeeshan Khan², Vineet Gandhi¹, Makarand Tapaswi^{1
  1}CVIT, IIIT Hyderabad, India
  ²Inria, Paris, France
• BIMBA: Selective-Scan Compression for Long-Range Video Question Answering
  Md Mohaiminul Islam¹, Tushar Nagarajan², Huiyu Wang², Gedas Bertasius¹, Lorenzo Torresani^{2
  1}UNC Chapel Hill
  ²Meta AI
• PG-Video-LLaVA: Pixel Grounding Large Video-Language Models
  Shehan Munasinghe¹, Rusiru Thushara¹, Muhammad Maaz¹, Hanoona Abdul Rasheed¹, Salman Khan^1,2, Mubarak Shah³, Fahad Shahbaz Khan^{1,4
  1}Mohamed bin Zayed University of AI
  ²Australian National University
  ³University of Central Florida
  ⁴Linköping University
• Contextual Self-paced Learning for Weakly Supervised Spatio-Temporal Video Grounding
  Akash Kumar¹, Zsolt Kira², Yogesh S Rawat^{2
  1}University of Central Florida
  ² Georgia Institute of Technology
• STPro: Spatial and Temporal Progressive Learning for Weakly Supervised Spatio-Temporal Grounding
  Aaryan Garg¹, Akash Kumar², Yogesh S Rawat^{2
  1}BITS Pilani
  ²CRCV, University of Central Florida
• Number it: Temporal Grounding Videos like Flipping Manga
  Yongliang Wu^1,2,4, Xinting Hu³, Yuyang Sun^1,2, Yizhou Zhou⁴, Wenbo Zhu⁵, Fengyun Rao⁴, Bernt Schiele³, Xu Yang^{1,2
  1}Southeast University, China
  ²Key Laboratory of New Generation Artificial Intelligence Technology & Its Interdisciplinary Applications (Southeast University), Ministry of Education, China
  ³Max Planck Institute for Informatics, Saarland Informatics Campus, Germany
  ⁴WeChat Vision, Tencent Inc., China
  ⁵University of California, Berkeley, America
• Lost in Time: A New Temporal Benchmark for Video LLMs
  Daniel Cores¹, Michael Dorkenwald², Manuel Mucientes¹, Cees G. M. Snoek², Yuki M Asano^{3
  1}CiTIUS, University of Santiago de Compostela
  ²QUVA Lab, University of Amsterdam
  ³Fundamental AI Lab, University of Technology Nuremberg
• ARGUS: Hallucination and Omission Evaluation in Video-LLMs
  Ruchit Rawal¹, Reza Shirkavand¹, Heng Huang¹, Gowthami Somepalli¹, Tom Goldstein^{1
  1}University of Maryland, College Park
• Moment Sampling in Video LLMs for Long-Form Video QA
  Mustafa Chasmai¹, Gauri Jagatap², Gouthaman KV², Grant Van Horn¹, Subhransu Maji¹, Andrea Fanelli^{2
  1}University of Massachusetts Amherst
  ²Dolby Laboratories