VideoINSTA: Zero-shot Long Video Understanding via Informative Spatial-Temporal Reasoning with LLMs
It addresses the problem of complex reasoning over extended timespans in long videos for researchers and practitioners in video-language AI, though it is incremental as it builds on existing zero-shot LLM approaches.
The paper tackles the challenge of zero-shot long video understanding by proposing VideoINSTA, a framework that uses informative spatial-temporal reasoning with LLMs, achieving significant improvements on benchmarks like EgoSchema, NextQA, IntentQA, and ActivityNetQA.
In the video-language domain, recent works in leveraging zero-shot Large Language Model-based reasoning for video understanding have become competitive challengers to previous end-to-end models. However, long video understanding presents unique challenges due to the complexity of reasoning over extended timespans, even for zero-shot LLM-based approaches. The challenge of information redundancy in long videos prompts the question of what specific information is essential for large language models (LLMs) and how to leverage them for complex spatial-temporal reasoning in long-form video analysis. We propose a framework VideoINSTA, i.e. INformative Spatial-TemporAl Reasoning for zero-shot long-form video understanding. VideoINSTA contributes (1) a zero-shot framework for long video understanding using LLMs; (2) an event-based temporal reasoning and content-based spatial reasoning approach for LLMs to reason over spatial-temporal information in videos; (3) a self-reflective information reasoning scheme balancing temporal factors based on information sufficiency and prediction confidence. Our model significantly improves the state-of-the-art on three long video question-answering benchmarks: EgoSchema, NextQA, and IntentQA, and the open question answering dataset ActivityNetQA. The code is released here: https://github.com/mayhugotong/VideoINSTA.