MiniGPT4-Video: Advancing Multimodal LLMs for Video Understanding with Interleaved Visual-Textual Tokens

TL;DR

MiniGPT4-Video is a multimodal LLM that processes video sequences and textual data, enabling advanced video understanding and question answering, outperforming existing methods on multiple benchmarks.

Contribution

It extends MiniGPT-v2 to handle temporal video data and multimodal interactions, significantly improving video understanding capabilities.

Findings

Outperforms state-of-the-art on MSVD, MSRVTT, TGIF, TVQA benchmarks

Achieves up to 20.82% improvement on TGIF

Successfully integrates visual and textual data for comprehensive video analysis

Abstract

This paper introduces MiniGPT4-Video, a multimodal Large Language Model (LLM) designed specifically for video understanding. The model is capable of processing both temporal visual and textual data, making it adept at understanding the complexities of videos. Building upon the success of MiniGPT-v2, which excelled in translating visual features into the LLM space for single images and achieved impressive results on various image-text benchmarks, this paper extends the model's capabilities to process a sequence of frames, enabling it to comprehend videos. MiniGPT4-video does not only consider visual content but also incorporates textual conversations, allowing the model to effectively answer queries involving both visual and text components. The proposed model outperforms existing state-of-the-art methods, registering gains of 4.22%, 1.13%, 20.82%, and 13.1% on the MSVD, MSRVTT, TGIF, and TVQA benchmarks respectively. Our models and code have been made publicly available here https://vision-cair.github.io/MiniGPT4-video/