Learning Robust Self-attention Features for Speech Emotion Recognition with Label-adaptive Mixup

Open in new window