Long Movie Clip Classification with State-Space Video Models

TL;DR

ViS4mer is an efficient long-range video model combining Transformer and S4 layers, enabling better long movie classification with less computation and memory, achieving state-of-the-art results on multiple benchmarks.

Contribution

The paper introduces ViS4mer, a novel model that integrates Transformer and structured state-space layers for scalable long-range video understanding.

Findings

ViS4mer is 2.63x faster than pure self-attention models.

It requires 8x less GPU memory.

Achieves state-of-the-art results on 6 out of 9 long video classification tasks.

Abstract

Most modern video recognition models are designed to operate on short video clips (e.g., 5-10s in length). Thus, it is challenging to apply such models to long movie understanding tasks, which typically require sophisticated long-range temporal reasoning. The recently introduced video transformers partially address this issue by using long-range temporal self-attention. However, due to the quadratic cost of self-attention, such models are often costly and impractical to use. Instead, we propose ViS4mer, an efficient long-range video model that combines the strengths of self-attention and the recently introduced structured state-space sequence (S4) layer. Our model uses a standard Transformer encoder for short-range spatiotemporal feature extraction, and a multi-scale temporal S4 decoder for subsequent long-range temporal reasoning. By progressively reducing the spatiotemporal feature resolution and channel dimension at each decoder layer, ViS4mer learns complex long-range spatiotemporal dependencies in a video. Furthermore, ViS4mer is $2.63\times$ faster and requires $8\times$ less GPU memory than the corresponding pure self-attention-based model. Additionally, ViS4mer achieves state-of-the-art results in $6$ out of $9$ long-form movie video classification tasks on the Long Video Understanding (LVU) benchmark. Furthermore, we show that our approach successfully generalizes to other domains, achieving competitive results on the Breakfast and the COIN procedural activity datasets. The code is publicly available at: https://github.com/md-mohaiminul/ViS4mer.