Speech SIMCLR: Combining Contrastive and Reconstruction Objective for Self-supervised Speech Representation Learning
This work addresses speech representation learning for tasks like emotion and speech recognition, but it is incremental as it adapts an existing visual method to speech.
The paper tackled self-supervised speech representation learning by proposing Speech SimCLR, which combines contrastive and reconstruction objectives, achieving competitive results on speech emotion recognition and speech recognition tasks.
Self-supervised visual pretraining has shown significant progress recently. Among those methods, SimCLR greatly advanced the state of the art in self-supervised and semi-supervised learning on ImageNet. The input feature representations for speech and visual tasks are both continuous, so it is natural to consider applying similar objective on speech representation learning. In this paper, we propose Speech SimCLR, a new self-supervised objective for speech representation learning. During training, Speech SimCLR applies augmentation on raw speech and its spectrogram. Its objective is the combination of contrastive loss that maximizes agreement between differently augmented samples in the latent space and reconstruction loss of input representation. The proposed method achieved competitive results on speech emotion recognition and speech recognition.