RevColV2

RevColV2 is a new architecture that learns disentangled representations during MIM pre-training and keeps a unified autoencoder architecture when transferring to downstream tasks.

Data and Resources

Cite this as

Qi Han, Yuxuan Cai, Xiangyu Zhang (2024). Dataset: RevColV2. https://doi.org/10.57702/j0wllluy

DOI retrieved: December 3, 2024

Additional Info

Field Value
Created December 3, 2024
Last update December 3, 2024
Author Qi Han
More Authors
Yuxuan Cai
Xiangyu Zhang
Homepage https://github.com/megvii-research/RevCol