Closed Dokumushikun closed 3 years ago
Hello. In your paper, you mentioned using negtive labels that are diffent from the true labels to train the network, but how can we get these true labels under the unsupervised situation?
Hello. In your paper, you mentioned using negtive labels that are diffent from the true labels to train the network, but how can we get these true labels under the unsupervised situation?