We introduce Three Towers (3T), a flexible method to improve the contrastive learning of vision-language models by incorporating pretrained image classifiers. While contrastive models are usually trained from scratch, LiT (Zhai et al., 2022) has recently shown performance gains from using pretrained classifier embeddings. However, LiT directly replaces the image tower with the frozen embeddings, excluding any potential benefits from training the image tower contrastively. With 3T, we propose a more flexible strategy that allows the image tower to benefit from both pretrained embeddings and contrastive training. To achieve this, we introduce a third tower that contains the frozen pretrained embeddings, and we encourage alignment between this third tower and the main image-text towers. Empirically, 3T consistently improves over LiT and the CLIP-style from-scratch baseline for retrieval tasks. For classification, 3T reliably improves over the from-scratch baseline, and while it underperforms relative to LiT for JFT-pretrained models, it outperforms LiT for ImageNet-21k and Places365 pretraining.
@article{arxiv.2305.16999,
title = {Three Towers: Flexible Contrastive Learning with Pretrained Image Models},
author = {Jannik Kossen and Mark Collier and Basil Mustafa and Xiao Wang and Xiaohua Zhai and Lucas Beyer and Andreas Steiner and Jesse Berent and Rodolphe Jenatton and Efi Kokiopoulou},
journal= {arXiv preprint arXiv:2305.16999},
year = {2023}
}