Search Results for author:
Found 2 papers, 1 papers with code
Vision Transformers with Mixed-Resolution Tokenization
In this work, we apply this approach to Vision Transformers by introducing a novel image tokenization scheme, replacing the standard uniform grid with a mixed-resolution sequence of tokens, where each token represents a patch of arbitrary size.
Diverse Imagenet Models Transfer Better
Encouraged by the recent transferability results of self-supervised models, we propose a method that combines self-supervised and supervised pretraining to generate models with both high diversity and high accuracy, and as a result high transferability.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
