CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noise
This addresses the scalability issue of label noise correction for image classification, offering a practical solution that reduces human effort while maintaining effectiveness.
The paper tackles the problem of training image classifiers with noisy labels by introducing CleanNet, a transfer learning approach that reduces the need for human verification. The method achieves a 41.5% reduction in label noise detection error on unseen classes and reaches 47% of the performance gain of full verification with only 3.2% of images verified.
In this paper, we study the problem of learning image classification models with label noise. Existing approaches depending on human supervision are generally not scalable as manually identifying correct or incorrect labels is time-consuming, whereas approaches not relying on human supervision are scalable but less effective. To reduce the amount of human supervision for label noise cleaning, we introduce CleanNet, a joint neural embedding network, which only requires a fraction of the classes being manually verified to provide the knowledge of label noise that can be transferred to other classes. We further integrate CleanNet and conventional convolutional neural network classifier into one framework for image classification learning. We demonstrate the effectiveness of the proposed algorithm on both of the label noise detection task and the image classification on noisy data task on several large-scale datasets. Experimental results show that CleanNet can reduce label noise detection error rate on held-out classes where no human supervision available by 41.5% compared to current weakly supervised methods. It also achieves 47% of the performance gain of verifying all images with only 3.2% images verified on an image classification task. Source code and dataset will be available at kuanghuei.github.io/CleanNetProject.