Look, Remember and Reason: Grounded reasoning in videos with language models

TL;DR

This paper introduces a grounded reasoning framework for videos using language models trained on low-level visual tasks, enabling better causal and compositional reasoning in videos.

Contribution

It proposes an end-to-end training approach that combines low-level visual skills with language models to improve video reasoning capabilities.

Findings

Outperforms state-of-the-art methods on multiple datasets

Effective in causal and compositional spatiotemporal reasoning

Utilizes a two-stream video encoder with spatiotemporal attention

Abstract

Multi-modal language models (LM) have recently shown promising performance in high-level reasoning tasks on videos. However, existing methods still fall short in tasks like causal or compositional spatiotemporal reasoning over actions, in which model predictions need to be grounded in fine-grained low-level details, such as object motions and object interactions. In this work, we propose training an LM end-to-end on low-level surrogate tasks, including object detection, re-identification, and tracking, to endow the model with the required low-level visual capabilities. We show that a two-stream video encoder with spatiotemporal attention is effective at capturing the required static and motion-based cues in the video. By leveraging the LM's ability to perform the low-level surrogate tasks, we can cast reasoning in videos as the three-step process of Look, Remember, Reason wherein visual information is extracted using low-level visual skills step-by-step and then integrated to arrive at a final answer. We demonstrate the effectiveness of our framework on diverse visual reasoning tasks from the ACRE, CATER, Something-Else and STAR datasets. Our approach is trainable end-to-end and surpasses state-of-the-art task-specific methods across these tasks by a large margin.