MutexMatch: Semi-Supervised Learning with Mutex-Based Consistency Regularization

27 Mar 2022  ยท  Yue Duan, Zhen Zhao, Lei Qi, Lei Wang, Luping Zhou, Yinghuan Shi, Yang Gao ยท

The core issue in semi-supervised learning (SSL) lies in how to effectively leverage unlabeled data, whereas most existing methods tend to put a great emphasis on the utilization of high-confidence samples yet seldom fully explore the usage of low-confidence samples. In this paper, we aim to utilize low-confidence samples in a novel way with our proposed mutex-based consistency regularization, namely MutexMatch. Specifically, the high-confidence samples are required to exactly predict "what it is" by conventional True-Positive Classifier, while the low-confidence samples are employed to achieve a simpler goal -- to predict with ease "what it is not" by True-Negative Classifier. In this sense, we not only mitigate the pseudo-labeling errors but also make full use of the low-confidence unlabeled data by consistency of dissimilarity degree. MutexMatch achieves superior performance on multiple benchmark datasets, i.e., CIFAR-10, CIFAR-100, SVHN, STL-10, mini-ImageNet and Tiny-ImageNet. More importantly, our method further shows superiority when the amount of labeled data is scarce, e.g., 92.23% accuracy with only 20 labeled data on CIFAR-10. Our code and model weights have been released at https://github.com/NJUyued/MutexMatch4SSL.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Semi-Supervised Image Classification CIFAR-100, 200 Labels MutexMatch (k=0.6C) Percentage error 58.41 # 1
Semi-Supervised Image Classification cifar-10, 10 Labels MutexMatch Accuracy (Test) 76.06 # 2
Semi-Supervised Image Classification CIFAR-10, 20 Labels MutexMatch (k=0.6C) Percentage error 7.77 # 1
Semi-Supervised Image Classification CIFAR-10, 40 Labels MutexMatch (k=0.6C) Percentage error 5.79 # 9
Semi-Supervised Image Classification CIFAR-10, 80 Labels MutexMatch (k=0.6C) Percentage error 5 # 1
Semi-Supervised Image Classification Mini-ImageNet, 1000 Labels MutexMatch Accuracy 48.04 # 1
Semi-Supervised Image Classification SVHN, 250 Labels MutexMatch (k=0.6C) Accuracy 97.47 # 4
Semi-Supervised Image Classification SVHN, 40 Labels MutexMatch (k=0.6C) Percentage error 3.45 # 3

Methods


No methods listed for this paper. Add relevant methods here