GraphVid: It Only Takes a Few Nodes to Understand a Video

TL;DR

GraphVid introduces a graph-based video representation using superpixels and GCNs, significantly reducing computational costs while maintaining competitive performance, enabling efficient video understanding on limited hardware.

Contribution

The paper presents a novel superpixel-based graph representation for videos processed with GCNs, achieving high efficiency and comparable accuracy with fewer resources.

Findings

Reduces computational requirements 10-fold

Maintains competitive accuracy on Kinetics-400 and Charades

Enables resource-limited hardware to perform video analysis

Abstract

We propose a concise representation of videos that encode perceptually meaningful features into graphs. With this representation, we aim to leverage the large amount of redundancies in videos and save computations. First, we construct superpixel-based graph representations of videos by considering superpixels as graph nodes and create spatial and temporal connections between adjacent superpixels. Then, we leverage Graph Convolutional Networks to process this representation and predict the desired output. As a result, we are able to train models with much fewer parameters, which translates into short training periods and a reduction in computation resource requirements. A comprehensive experimental study on the publicly available datasets Kinetics-400 and Charades shows that the proposed method is highly cost-effective and uses limited commodity hardware during training and inference. It reduces the computational requirements 10-fold while achieving results that are comparable to state-of-the-art methods. We believe that the proposed approach is a promising direction that could open the door to solving video understanding more efficiently and enable more resource limited users to thrive in this research field.