Masked Autoencoders Are Scalable Vision Learners

Code

• github.com/facebookresearch/mae
  Officialpytorch★ 8,245
• github.com/lightly-ai/lightly
  pytorch★ 3,700
• github.com/open-mmlab/mmselfsup
  pytorch★ 3,297
• github.com/pengzhiliang/MAE-pytorch
  pytorch★ 2,682
• github.com/alibaba/EasyCV
  pytorch★ 1,949
• github.com/facebookresearch/multimodal
  pytorch★ 1,706
• github.com/facebookresearch/hiera
  pytorch★ 1,059
• github.com/Westlake-AI/openmixup
  pytorch★ 657
• github.com/oneflow-inc/libai
  none★ 406
• github.com/xplip/pixel
  pytorch★ 346

Abstract

This paper shows that masked autoencoders (MAE) are scalable self-supervised learners for computer vision. Our MAE approach is simple: we mask random patches of the input image and reconstruct the missing pixels. It is based on two core designs. First, we develop an asymmetric encoder-decoder architecture, with an encoder that operates only on the visible subset of patches (without mask tokens), along with a lightweight decoder that reconstructs the original image from the latent representation and mask tokens. Second, we find that masking a high proportion of the input image, e.g., 75%, yields a nontrivial and meaningful self-supervisory task. Coupling these two designs enables us to train large models efficiently and effectively: we accelerate training (by 3x or more) and improve accuracy. Our scalable approach allows for learning high-capacity models that generalize well: e.g., a vanilla ViT-Huge model achieves the best accuracy (87.8%) among methods that use only ImageNet-1K data. Transfer performance in downstream tasks outperforms supervised pre-training and shows promising scaling behavior.

Tasks

Benchmark Results

Reproductions