Recurrent Attention-based Token Selection for Efficient Streaming Video-LLMs

TL;DR

This paper introduces a training-free, attention-based token selection method for streaming Video-LLMs that significantly reduces visual token processing while maintaining high performance in real-time video understanding.

Contribution

It presents a novel, training-free approach that uses LLM-informed token selection and recurrent processing to enable efficient streaming video analysis with minimal performance loss.

Findings

Discards up to 95% of unimportant visual tokens with minimal performance impact

Achieves state-of-the-art results on streaming video benchmarks

Balances efficiency and effectiveness in real-time video understanding

Abstract

Video Large Language Models (Video-LLMs) excel at understanding videos in-context, provided they have full access to the video when answering queries. However, these models face challenges in streaming scenarios where hour-long videos must be processed online, and questions need timely responses. In this work, we propose a training-free approach compatible with standard Video-LLMs, leveraging three key concepts: 1) LLM-informed selection of visual tokens to identify those that the LLM has attended to and contributed to its understanding of each short clip. Our attention-based selection allows us to discard up to ~95% of unimportant visual tokens with minimal performance loss; 2) Recurrent processing of past selected tokens to generate temporally coherent understanding of each processed clip; 3) Caption-based question answering for lightweight and accurate responses. Our method achieves state-of-the-art performance on streaming video benchmarks, striking a balance between efficiency and effectiveness.