M-MAE
PublicOfficial implementation of Matrix Variational Masked Autoencoder (M-MAE) for ICML paper "Information Flow in Self-Supervised Learning" (https://arxiv.org/abs/2309.17281)
foundation-modelsmasked-autoencodermasked-image-modelingnon-contrastive-learningpretrainingself-supervised-learning
Creat：2023-09-27T11:10:31
Update：2024-11-25T05:37:24
https://arxiv.org/abs/2309.17281
14
Stars
0
Stars Increase