A Unified Mixture-View Framework for Unsupervised Representation Learning
Xiangxiang Chu (Meituan), Xiaohang Zhan (The Chinese University of Hong Kong), Bo Zhang (Meituan)*
The 33rd British Machine Vision Conference
Abstract
Recent unsupervised contrastive representation learning follows a Single Instance Multi-view (SIM) paradigm where positive pairs are usually constructed with intra-image data augmentation. In this paper, we propose an effective approach called Beyond Single Instance Multi-view (BSIM). Specifically, we impose more accurate instance discrimination capability by measuring the joint similarity between two randomly sampled instances and their mixture, namely spurious-positive pairs. We believe that learning joint similarity helps to improve the performance when encoded features are distributed more evenly in the latent space. We apply it as an orthogonal improvement for unsupervised contrastive representation learning, including current outstanding methods SimCLR, MoCo, BYOL and SimSiam. We evaluate our learned representations on many downstream benchmarks like linear classification on ImageNet-1k and PASCAL VOC 2007, object detection on MS COCO 2017 and VOC, etc. We obtain substantial gains with a large margin almost on all these tasks compared with prior arts.Video
Citation
@inproceedings{Chu_2022_BMVC,
author = {Xiangxiang Chu and Xiaohang Zhan and Bo Zhang},
title = {A Unified Mixture-View Framework for Unsupervised Representation Learning},
booktitle = {33rd British Machine Vision Conference 2022, {BMVC} 2022, London, UK, November 21-24, 2022},
publisher = {{BMVA} Press},
year = {2022},
url = {https://bmvc2022.mpi-inf.mpg.de/0447.pdf}
}
Copyright © 2022
The British Machine Vision Conference is organised by