Visual Question Answering

Also known as VQA

The field asks how machines can connect a question to relevant visual evidence, read text and diagrams, and combine perception with commonsense or specialized knowledge. Approaches range from vision-language encoders and multimodal large language models to region and crop grounding, OCR and scene graphs, retrieval-augmented agents, and fine-tuning or reinforcement learning. Benchmarks cover everyday images and compositional reasoning, knowledge-based and text-rich questions, and document, video, medical, scientific, and multi-view settings. A central challenge is that correct-sounding answers may lack reliable visual grounding: models still miss small details and spatial relations, struggle to combine evidence across views or long documents, and may fail to recognize when they should abstain.

Latest papers

253 papers36 papers in the last four weeks, up 13% on the four weeks before. 0.2% of all new papers.

Rethinking Token Reduction for Large Vision-Language Models
Mar 23, 2026Yi Wang, Haofei Zhang, Qihan Huang +7Efficient VLM Inference Visual Token Compression
College of Computer Science and Technology, Zhejiang University · State Key Laboratory of Blockchain and Data Security, Zhejiang University · Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security +3
Hidden Clones: Exposing and Fixing Family Bias in Vision-Language Model Ensembles
Mar 17, 2026Zacharie BugaudVision-Language Models Visual Question Answering
Astera Institute
360° Image Perception with MLLMs: A Comprehensive Benchmark and a Training-Free Method
Mar 17, 2026Huyen T. T. Tran, Van-Quang Nguyen, Farros Alferro +2Visual Question Answering Visual Spatial Reasoning
GSIS, Tohoku University · RIKEN AIP, Japan
Visual Set Program Synthesizer
Mar 16, 2026Zehua Cheng, Wei Dai, Wenhu Zhang +2Visual Program Synthesis Visual Reasoning
Department of Computer Science, University of Oxford · Department of Computer Science and Engineering, Hong Kong University of Science and Technology · Institute of Logic and Computation, TU Wien
SEA-Vision: A Multilingual Benchmark for Comprehensive Document and Scene Text Understanding in Southeast Asia
Mar 16, 2026Pengfei Yue, Xingran Zhao, Juntao Chen +6Multilingual VLM Evaluation Document Parsing
Xiamen University, China · Shopee, China · Tongji University, China
LLMind: Bio-inspired Training-free Adaptive Visual Representations for Vision-Language Models
Mar 16, 2026Soumyaratna Debnath, Bui Duc Manh, Zinan Liu +1Foveated Vision Efficient VLM Inference
Nanyang Technological University, Singapore
AgroOmni: A Large-Scale Multi-view Agricultural Dataset for Cross-Scale Multimodal Reasoning
Mar 15, 2026Jiarui Zhang, Junqi Hu, Zurong Mai +10Multimodal Reasoning Multimodal Large Language Models
Sun Yat-sen University · The Chinese University of Hong Kong · HuanTian Wisdom Technology Co., Ltd. +4
Deeper Thought, Weaker Aim: Understanding and Mitigating Perceptual Impairment during Reasoning in Multimodal Large Language Models
Mar 15, 2026Ruiying Peng, Xueyu Wu, Jing Lei +3Multimodal Large Language Models VLM Reasoning
Tsinghua Shenzhen International Graduate School · Huawei Technologies
AgriChat: A Multimodal Large Language Model for Agriculture Image Understanding
Mar 14, 2026Abderrahmene Boudiaf, Irfan Hussain, Sajid JavedAgricultural Image Analysis Vision-Language Models
Department of Computer Science, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates
MaterialFigBENCH: benchmark dataset with figures for evaluating college-level materials science problem-solving abilities of multimodal large language models
Mar 12, 2026Michiko Yoshitake, Yuta Suzuki, Ryo Igarashi +2Multimodal Model Evaluation Multimodal Large Language Models
OMRON SINIC X, Tokyo, Japan · Department of Applied Physics, Osaka Univ, Suita, Japan · Department of Mechanical Engineering, Univ. Tokyo, Tokyo, Japan
Taking Shortcuts for Categorical VQA Using Super Neurons
Mar 11, 2026Pierre Musacchio, Jaeyi Jeong, Dahun Kim +1Efficient VLM Inference Visual Question Answering
Seoul National University · EPFL · Google Deepmind
AutoViVQA: A Large-Scale Automatically Constructed Dataset for Vietnamese Visual Question Answering
Mar 10, 2026Nguyen Anh Tuong, Phan Ba Duc, Nguyen Trung Quoc +4Visual Question Answering Multimodal QA
Faculty of Information Technology, University of Science, VNU-HCM, Vietnam · Vietnam National University, Ho Chi Minh City, Vietnam
More than the Sum: Panorama-Language Models for Adverse Omni-Scenes
Mar 10, 2026Weijia Fan, Ruiping Liu, Jiale Wei +7Vision-Language Models Visual Question Answering
Karlsruhe Institute of Technology · Shenzhen University · UCL +1
Hospitality-VQA: Decision-Oriented Informativeness Evaluation for Vision-Language Models
Mar 9, 2026Jeongwoo Lee, Baek Duhyeong, Eungyeol Han +5VLM Evaluation Visual Question Answering
Yonsei University · Yanolja NEXT
AQuA: Toward Strategic Response Generation for Ambiguous Visual Questions
Mar 8, 2026Jihyoung Jang, Hyounghun KimVisual Question Answering VLM Adaptation
Graduate School of Artificial Intelligence, POSTECH · Department of Computer Science and Engineering, POSTECH
PatchCue: Enhancing Vision-Language Model Reasoning with Patch-Based Visual Cues
Mar 6, 2026Yukun Qi, Pei Fu, Hang Li +5VLM Reasoning Visual Reasoning
MiLM Plus, Xiaomi Inc.
Wiki-R1: Incentivizing Multimodal Reasoning for Knowledge-based VQA via Data and Sampling Curriculum
Mar 5, 2026Shan Ning, Longtian Qiu, Xuming HeVisual Question Answering Reinforcement Learning
ShanghaiTech University, Shanghai, China · Lingang Laboratory, Shanghai, China · Shanghai Engineering Research Center of Intelligent Vision and Imaging
Through the Lens of Contrast: Self-Improving Visual Reasoning in VLMs
Mar 3, 2026Zhiyu Pan, Yizheng Wu, Jiashen Hua +5VLM Reasoning VLM Hallucination Mitigation
Huazhong University of Science and Technology · Alibaba Cloud
ORCA: Orchestrated Reasoning with Collaborative Agents for Document Visual Question Answering
Mar 2, 2026Aymen Lassoued, Mohamed Ali Souibgui, Yousri KessentiniVisual Question Answering Vision-Language Agents
Digital Research Center of Sfax, SMARTS Laboratory, Sfax, Tunisia. · ´Ecole Polytechnique de Tunisie, University of Carthage · Computer Vision Center, Universitat Aut`onoma de Barcelona
SUPERGLASSES: Benchmarking Vision Language Models as Intelligent Agents for AI Smart Glasses
Feb 26, 2026Zhuohang Jiang, Xu Yuan, Haohao Qu +4Visual Question Answering Knowledge-Based VQA
The Hong Kong Polytechnic University
ConFoThinking: Consolidated Focused Attention Driven Thinking for Visual Question Answering
Feb 26, 2026Zhaodong Wu, Haochen Xue, Qi Cao +5Visual Question Answering VLM Reasoning
Wangxuan Institute of Computer Technology, Peking University, Beijing, China · School of AI and Advanced Computing, Xi’an Jiaotong-Liverpool University, Suzhou, China
SimpleOCR: Rendering Visualized Questions to Teach MLLMs to Read
Feb 25, 2026Yibo Peng, Peng Xia, Ding Zhong +6Multimodal Large Language Models Optical Character Recognition
UNC-Chapel Hill · Carnegie Mellon University · University of Michigan +1
PanoEnv: Exploring 3D Spatial Intelligence in Panoramic Environments with Reinforcement Learning
Feb 25, 2026Zekai Lin, Xu Zheng3D Spatial Reasoning 3D VQA
University of Glasgow · HKUST(GZ)
UDVideoQA: A Traffic Video Question Answering Dataset for Multi-Object Spatio-Temporal Reasoning in Urban Dynamics
Feb 24, 2026Joseph Raj Vishal, Nagasiri Poluri, Katha Naik +7Video QA Visual Question Answering
Arizona State University
PositionOCR: Augmenting Positional Awareness in Multi-Modal Models via Hybrid Specialist Integration
Feb 22, 2026Chen Duan, Zhentao Guo, Pei Fu +3Multimodal Large Language Models Optical Character Recognition
Say It My Way: Exploring Control in Conversational Visual Question Answering with Blind Users
Feb 18, 2026Farnaz Zamiri Zeraati, Yang Trista Cao, Yuehan Qiao +2Assistive Technology for Blind and Low-Vision People Human-AI Interaction
University of Maryland, College Park, Maryland, USA · University of Texas at Austin, Austin, Texas, USA
IRIS: Intent Resolution via Inference-time Saccades for Open-Ended VQA in Large Vision-Language Models
Feb 18, 2026Parsa Madinei, Srijita Karmakar, Russell Cohen Hoffing +2Visual Question Answering Large Vision-Language Models
Department of Computer Science, UC Santa Barbara · Department of Psychological & Brain Sciences, UC Santa Barbara · DEVCOM Army Research Laboratory
MaS-VQA: A Mask-and-Select Framework for Knowledge-Based Visual Question Answering
Feb 17, 2026Xianwei Mao, Kai Ye, Sheng Zhou +4Visual Question Answering Knowledge-Based VQA
Zhejiang University, Hangzhou, China · Alibaba Group, Hangzhou, China
How to Train Your Long-Context Visual Document Model
Feb 16, 2026Austin VeselkaVisual Question Answering Vision-Language Models
LightOn
ThermEval: A Structured Benchmark for Evaluation of Vision-Language Models on Thermal Imagery
Feb 16, 2026Ayush Shrivastava, Kirtan Gangani, Laksh Jain +2VLM Evaluation Thermal Imaging
Indian Institute of Technology, Gandhinagar, India · Carnegie Mellon University, Pittsburgh, USA

Mar 23, 2026J/K move · Enter open · S save

Rethinking Token Reduction for Large Vision-Language Models

Yi Wang, Haofei Zhang, Qihan Huang, Anda Cao+6

College of Computer Science and Technology, Zhejiang University · State Key Laboratory of Blockchain and Data Security, Zhejiang University · Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security+3

Efficient VLM Inference Visual Token Compression Large Vision-Language Models Visual Token Pruning Visual Question Answering

Open paper AskPDF arXiv

Large Vision-Language Models (LVLMs) excel in visual understanding and reasoning, but the excessive visual tokens lead to high inference costs. Although recent token reduction methods mitigate this issue, they mainly target single-turn Visual Question Answering (VQA), leaving the more practical multi-turn VQA (MT-VQA) scenario largely unexplored. MT-VQA introduces additional challenges, as subsequent questions are unknown beforehand and may refer to arbitrary image regions, making existing reduction strategies ineffective. Specifically, current approaches fall into two categories: prompt-dependent methods, which bias toward the initial text prompt and discard information useful for subsequent turns; prompt-agnostic ones, which, though technically applicable to multi-turn settings, rely on heuristic reduction metrics such as attention scores, leading to suboptimal performance. In this paper, we propose a learning-based prompt-agnostic method, termed MetaCompress, overcoming the limitations of heuristic designs. We begin by formulating token reduction as a learnable compression mapping, unifying existing formats such as pruning and merging into a single learning objective. Upon this formulation, we introduce a data-efficient training paradigm capable of learning optimal compression mappings with limited computational costs. Extensive experiments on MT-VQA benchmarks and across multiple LVLM architectures demonstrate that MetaCompress achieves superior efficiency-accuracy trade-offs while maintaining strong generalization across dialogue turns. Our code is available at https://github.com/MArSha1147/MetaCompress.