Efficient Vision-Language Pre-Training by Cluster Masking
Abstract
We propose a simple strategy for masking image patches during visual-language contrastive learning that improves the quality of the learned representations and the training speed. During each iteration of training we randomly mask clusters of visually similar image patches as measured by their raw pixel intensities. This provides an extra learning signal beyond the contrastive training itself since it forces a model to predict words for masked visual structures solely from context. It also speeds up training by reducing the amount of data used in each image. We evaluate the effectiveness of our model by pre-training on a number of benchmarks finding that it outperforms other masking strategies such as FLIP on the quality of the learned representation.
Cite
Text
Wei et al. "Efficient Vision-Language Pre-Training by Cluster Masking." Conference on Computer Vision and Pattern Recognition, 2024. doi:10.1109/CVPR52733.2024.02532Markdown
[Wei et al. "Efficient Vision-Language Pre-Training by Cluster Masking." Conference on Computer Vision and Pattern Recognition, 2024.](https://mlanthology.org/cvpr/2024/wei2024cvpr-efficient/) doi:10.1109/CVPR52733.2024.02532BibTeX
@inproceedings{wei2024cvpr-efficient,
title = {{Efficient Vision-Language Pre-Training by Cluster Masking}},
author = {Wei, Zihao and Pan, Zixuan and Owens, Andrew},
booktitle = {Conference on Computer Vision and Pattern Recognition},
year = {2024},
pages = {26815-26825},
doi = {10.1109/CVPR52733.2024.02532},
url = {https://mlanthology.org/cvpr/2024/wei2024cvpr-efficient/}
}