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

作者：Chaitanya K. Ryali, Yuan-Ting Hu, Daniel Bolya, Wei Chen, Haoqi Fan, Po-Yao Huang, Vaibhav Aggarwal, Arkabandhu Chowdhury, Omid Poursaeed, Judy Hoffman, Jitendra Malik, Yanghao Li, Christoph Feichtenhofer · 发表于：arXiv (Cornell University) · 年份：2023 · DOI：10.48550/arxiv.2306.00989 · 被引用次数：61 · 研究领域：CCD and CMOS Imaging Sensors、Advanced Memory and Neural Computing、Advanced Neural Network Applications

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.