Agentic Very Long Video Understanding

TL;DR

This paper introduces EGAgent, a novel framework for long-horizon egocentric video understanding that leverages entity scene graphs and structured reasoning to interpret continuous, multi-modal video streams over days or weeks.

Contribution

EGAgent advances long-term video understanding by integrating entity scene graphs with planning and reasoning tools, enabling detailed cross-modal analysis of extended video streams.

Findings

Achieves state-of-the-art 57.5% on EgoLifeQA dataset.

Performs competitively with 74.1% on Video-MME (Long).

Demonstrates effective long-horizon, multi-modal reasoning.

Abstract

The advent of always-on personal AI assistants, enabled by all-day wearable devices such as smart glasses, demands a new level of contextual understanding, one that goes beyond short, isolated events to encompass the continuous, longitudinal stream of egocentric video. Achieving this vision requires advances in long-horizon video understanding, where systems must interpret and recall visual and audio information spanning days or even weeks. Existing methods, including large language models and retrieval-augmented generation, are constrained by limited context windows and lack the ability to perform compositional, multi-hop reasoning over very long video streams. In this work, we address these challenges through EGAgent, an enhanced agentic framework centered on entity scene graphs, which represent people, places, objects, and their relationships over time. Our system equips a planning agent with tools for structured search and reasoning over these graphs, as well as hybrid visual and audio search capabilities, enabling detailed, cross-modal, and temporally coherent reasoning. Experiments on the EgoLifeQA and Video-MME (Long) datasets show that our method achieves state-of-the-art performance on EgoLifeQA (57.5%) and competitive performance on Video-MME (Long) (74.1%) for complex longitudinal video understanding tasks. Code is available at https://github.com/facebookresearch/egagent.