Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles

TL;DR

Hiera is a simple, fast, and accurate hierarchical vision transformer that removes unnecessary components from complex models by leveraging strong pretraining, achieving superior performance with less complexity.

Contribution

The paper introduces Hiera, a streamlined hierarchical vision transformer that maintains high accuracy while significantly reducing complexity and computational costs, unlike previous models.

Findings

Hiera outperforms previous models in accuracy.

Hiera is faster at inference and training.

Hiera's simplicity does not compromise performance.

Abstract

Modern hierarchical vision transformers have added several vision-specific components in the pursuit of supervised classification performance. While these components lead to effective accuracies and attractive FLOP counts, the added complexity actually makes these transformers slower than their vanilla ViT counterparts. In this paper, we argue that this additional bulk is unnecessary. By pretraining with a strong visual pretext task (MAE), we can strip out all the bells-and-whistles from a state-of-the-art multi-stage vision transformer without losing accuracy. In the process, we create Hiera, an extremely simple hierarchical vision transformer that is more accurate than previous models while being significantly faster both at inference and during training. We evaluate Hiera on a variety of tasks for image and video recognition. Our code and models are available at https://github.com/facebookresearch/hiera.