Beyond the Single Camera: Agentic Multi-View Reasoning in Sports Video Understanding
2026-07-13 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors created SportMV-Bench, a new test for checking how well AI models understand sports videos taken from multiple camera angles. Sports videos are tricky because of fast action and blocked views, so using several viewpoints helps. They found that current AI models struggle to use these multiple views effectively, mainly because they have trouble spotting details and choosing the right views to look at. To fix this, the authors developed SportMV-Agent, which improves performance by actively picking views and carefully analyzing them, leading to better results.
Multimodal Large Language Models (MLLMs)multi-view video understandingsports video analysisvideo perceptionview selectionquestion-answering benchmarkevent interpretationactive perceptionevidence-grounded reasoningagentic framework
Authors
Kerui Chen, Jinglu Wang, Xiaoyi Zhang, Yan Lu
Abstract
Recent Multimodal Large Language Models (MLLMs) achieve strong performance on single-view video understanding benchmarks. However, sports videos involve dense occlusion, rapid motion, and complex interactions that are difficult to resolve from a single viewpoint. In practice, sports events are recorded from multiple camera angles, providing complementary evidence used by referees. Yet, no existing benchmark evaluates MLLMs on multi-view sports video understanding. To address this gap, we introduce SportMV-Bench, a comprehensive benchmark built from official match recordings, through a dedicated pipeline combining LLM-based generation, MLLM-based verification, and human filtering to ensure quality and consistency. SportMV-Bench containing 787 multi-view video bundles and 2592 question-answer pairs across three categories: Perception-Aware Recognition (PAR), Rule-aware Event Interpretation (REI), and Adjudicative Decision Reasoning(ADR). Our analysis shows that current MLLMs fail to effectively exploit multi-view information, with the bottlenecks lying in fine-grained visual perception and view selection rather than logical reasoning or domain knowledge. We propose SportMV-Agent, an agentic framework that orchestrates an iterative loop of active view selection, perception tool execution, and evidence-grounded reasoning, achieving a significant 14.46% relative improvement over the strongest MLLM baseline.