pengzhiliang/MAE-pytorch
PyTorch implementation of Masked Autoencoders (MAE), a self-supervised vision transformer model for scalable visual learning.
Not currently ranked — collecting fresh signals.
star history
This repository provides an unofficial PyTorch implementation of the MAE (Masked Autoencoders Are Scalable Vision Learners) paper. It implements the pretraining and fine-tuning processes for a vision transformer that learns by masking and reconstructing patches from images. The implementation includes shuffle/unshuffle operations for masked token handling and supports sine-cosine positional embeddings as described in the original paper.
Frequently asked
- What is pengzhiliang/MAE-pytorch?
- PyTorch implementation of Masked Autoencoders (MAE), a self-supervised vision transformer model for scalable visual learning.
- Is MAE-pytorch open source?
- Yes — pengzhiliang/MAE-pytorch is an open-source project tracked on heatdrop.
- What language is MAE-pytorch written in?
- pengzhiliang/MAE-pytorch is primarily written in Python.
- How popular is MAE-pytorch?
- pengzhiliang/MAE-pytorch has 2.7k stars on GitHub.
- Where can I find MAE-pytorch?
- pengzhiliang/MAE-pytorch is on GitHub at https://github.com/pengzhiliang/MAE-pytorch.