Weakly Supervised Contrastive Learning

Unsupervised visual representation learning has gained much attention from the computer vision community because of the recent achievement of contrastive learning. Most of the existing contrastive learning frameworks adopt the instance discrimination as the pretext task, which treating every single instance as a different class. However, such method will inevitably cause class collision problems, which hurts the quality of the learned representation. Motivated by this observation, we introduced a weakly supervised contrastive learning framework (WCL) to tackle this issue. Specifically, our proposed framework is based on two projection heads, one of which will perform the regular instance discrimination task. The other head will use a graph-based method to explore similar samples and generate a weak label, then perform a supervised contrastive learning task based on the weak label to pull the similar images closer. We further introduced a K-Nearest Neighbor based multi-crop strategy to expand the number of positive samples. Extensive experimental results demonstrate WCL improves the quality of self-supervised representations across different datasets. Notably, we get a new state-of-the-art result for semi-supervised learning. With only 1\% and 10\% labeled examples, WCL achieves 65\% and 72\% ImageNet Top-1 Accuracy using ResNet50, which is even higher than SimCLRv2 with ResNet101.

PDF Abstract ICCV 2021 PDF ICCV 2021 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Self-Supervised Image Classification ImageNet WCL (ResNet-50) Top 1 Accuracy 74.7% # 54
Number of Params 24M # 43
Semi-Supervised Image Classification ImageNet - 10% labeled data WCL (ResNet-50) Top 5 Accuracy 91.2% # 14
Top 1 Accuracy 72.0% # 21
Semi-Supervised Image Classification ImageNet - 1% labeled data WCL (ResNet-50) Top 5 Accuracy 86.3% # 11
Top 1 Accuracy 65.0% # 16