Visual Counting

Follow this topic Ask about this topic

Latest papers

21 papers6 papers in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

PhysCount: Do MLLMs Count Visual Appearances or Physical Entities?
Oct 29, 2026Lidong Lu, Jiaxin Lv, Liangrui Dong +5Visual Counting Object Counting
Nanjing University · NVIDIA
What do MLLMs know about 'Waldo'? Evaluating Visual Reasoning under Multi-Instance Multi-Object Challenges
Oct 28, 2026Saumya Mundra, Ajoy Mondal, Vineeth N Balasubramanian +1Visual Counting MLLM Evaluation
IIIT Hyderabad · Microsoft Research India · IIT Hyderabad
DecoyCount: Evaluating Visual-Reasoning Misalignment in Counting
Oct 26, 2026Qiuyi, Zhang, Zhen Xu +14Visual Counting Counting in VLMs
Richard · Elorian AI · Mercor
CountVideoBench: A Benchmark to Evaluate Prediction Biases on the Object Counting task by Video-Language Models
Oct 20, 2026Marc Chapus, Carlos Crispim-Junior, Véronique Eglin +1Counting in VLMs Visual Counting
INSA Lyon, Université Lumière Lyon 2, CNRS, Université Claude Bernard Lyon 1, LIRIS, UMR5205, 69621 Villeurbanne, France
Give the Model a Finger: Developmentally Aligned Evaluation of Counting in Vision-Language Models
Oct 20, 2026Yingxin Jia, Hui Geng, Wenxin Yang +1Counting in VLMs Vision-Language Models
Independent Researcher
UniCounting: Instance-Aware Proposal Consolidation for Image-Query-Free Multi-Category Counting
Oct 6, 2026Jinshi Liu, Pan Liu, Lei He +2Object Counting Data Association
Shenzhen University · The Hong Kong University of Science and Technology (Guangzhou) · Hunan University of Science and Technology +2
When Should the Count Change? Learning State Maintenance for Causal Video Counting
Sep 28, 2026Pengyiang Liu, Dongyue Lyu, Junbo Niu +3Streaming Video Understanding State Tracking
Beihang University · Peking University
NumBench: Diagnosing Counting Failures in Text-to-Image Models
Aug 28, 2026Sandeep Wadhwa, Mayank Vatsa, Richa Singh +2T2I Generation T2I Generation Evaluation
Indian Institute of Technology Jodhpur, India · Shiv Nadar University, India
Object Counting Across Modalities: Taxonomies, Benchmarks, Applications, and Open Challenges
Aug 24, 2026Joana Konadu Owusu, Shivanand Venkanna SheshappanavarObject Counting Visual Counting
Geometric Intelligence Research Lab., Dept. of Electrical Engineering and Computer Science, University of Wyoming, Laramie, WY, USA.
RECOUNT: Reference-guided Counting with Synthetic Visual Exemplars
Aug 20, 2026Adriano D'Alessandro, Ali Mahdavi-Amiri, Ghassan HamarnehObject Counting Visual Counting
Simon Fraser University
HoloCount: A Holistic Visual Counting Benchmark for MLLMs
Jul 7, 2026Jinhong Deng, Limeng Qiao, Guanglu WanMultimodal Robustness Object Counting
Meituan
ROSE: Benchmarking the Perception-to-Action Gap in Multimodal Models
Jun 18, 2026Yihao Wang, Zijian He, Jie Ren +1Multimodal Grounding Visual Counting
Sun Yat-sen University · Shaanxi Normal University
Count Anything
May 29, 2026Mengqi Lei, Shuokun Cheng, Wei Bao +4Object Counting Visual Counting
1{BNRist, THUIBCS, BLBCI, School of Software}, Tsinghua University · School of Computer Science and Technology, China University of Geosciences, Wuhan · 3State Key Laboratory of Human-Machine Hybrid Augmented Intelligence, National Engineering Research Center for Visual Information and Applications, and Institute of Artificial Intelligence and Robotics, Xi’an Jiaotong University
Unveiling the Visual Counting Bottleneck in Vision-Language Models
May 28, 2026Xingzhou Pang, Yifan Hou, Junling Wang +1VLM Reasoning Visual Counting
Department of Computer Science, ETH Zürich
Count Anything at Any Granularity
May 11, 2026Chang Liu, Haoning Wu, Weidi XieObject Counting Multimodal Large Language Models
School of Artificial Intelligence, Shanghai Jiao Tong University, China · CMIC, Shanghai Jiao Tong University, China
PushupBench: Your VLM is not good at counting pushups
Apr 25, 2026Shengzhi Li, Jiarun Chen, Karun Sharma +2Temporal Video Understanding Video-Language Model Evaluation
Counting to Four is still a Chore for VLMs
Apr 11, 2026Duy Le Dinh Anh, Patrick Amadeus Irawan, Tuan Van VoCounting in VLMs Attention Intervention in VLMs
MBZUAI
Getting to the Point: Pointing Improves LVLMs at Counting
Mar 23, 2026Simone Alghisi, Massimo Rizzoli, Seyed Mahed Mousavi +1Visual Counting Counting in VLMs
Signals and Interactive Systems Lab, University of Trento, Italy
Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Models
Mar 19, 2026Liwei Che, Zhiyu Xue, Yihao Quan +7Visual Counting Counting in VLMs
Rutgers University · UC Santa Barbara · University of Washington +1
UNICBench: UNIfied Counting Benchmark for MLLM
Feb 28, 2026Chenggang Rong, Tao Han, Zhiyuan Zhao +5Multimodal Large Language Models Multimodal Dataset
Northwestern Polytechnical University · Hong Kong University of Science and Technology · Institute of Artificial Intelligence (TeleAI), China Telecom +2
CountEx: Fine-Grained Counting via Exemplars and Exclusion
Feb 23, 2026Yifeng Huang, Gia Khanh Nguyen, Minh HoaiObject Counting Visual Counting
Department of Computer Science, Stony Brook University, Stony Brook, NY, USA · Australian Institute for Machine Learning, Adelaide University, SA, Australia

Oct 29, 2026J/K move · Enter open · S save

PhysCount: Do MLLMs Count Visual Appearances or Physical Entities?

Lidong Lu, Jiaxin Lv, Liangrui Dong, Xinyi Mao+4

Nanjing University · NVIDIA

Visual Counting Object Counting MLLM Evaluation

Open paper AskPDF arXiv

Images reveal appearances, but counting often asks about physical objects. A single entity may be visible directly, in several reflections, or through glass, so detecting every visible occurrence does not determine how many entities are present. Existing visual counting benchmarks rarely separate these two operations. We introduce PhysCount, a benchmark for evaluating whether multimodal large language models (MLLMs) can distinguish what appears in an image from what physically exists in the scene. PhysCount contains 546 real-world images and 4,595 human-written questions organized into 651 target-specific groups. For each group, annotators locate the relevant appearances, label how each is observed, and identify appearances of the same entity. Four complementary question families test the enumeration of appearances, the enumeration of physical entities, image-formation-specific counting, and appearance--entity correspondence. We measure both accuracy on each family and the proportion of groups answered entirely correctly. Across 40 MLLMs, counting visible appearances is substantially easier than recovering the identities that connect them. Even the strongest model answers every question correctly in only 13.36% of groups. Performance declines when one entity produces many appearances or when its appearances cross observation modes. Higher image resolution improves access to visual evidence but brings limited gains in complete-group accuracy, whereas answering related questions jointly produces a marked improvement. These results expose a gap between recognizing what is visible and correctly recovering the physical entities behind those observations.