Improved Speech Representations with Multi-Target Autoregressive Predictive Coding
This work addresses the challenge of improving speech representation learning for downstream applications, representing an incremental advancement over existing predictive coding methods.
The paper tackles the problem of learning meaningful speech representations from unlabeled data by extending Autoregressive Predictive Coding with an auxiliary objective to improve future frame prediction, resulting in richer phonetic content and enhanced performance on tasks like phonetic classification, speech recognition, and speech translation.
Training objectives based on predictive coding have recently been shown to be very effective at learning meaningful representations from unlabeled speech. One example is Autoregressive Predictive Coding (Chung et al., 2019), which trains an autoregressive RNN to generate an unseen future frame given a context such as recent past frames. The basic hypothesis of these approaches is that hidden states that can accurately predict future frames are a useful representation for many downstream tasks. In this paper we extend this hypothesis and aim to enrich the information encoded in the hidden states by training the model to make more accurate future predictions. We propose an auxiliary objective that serves as a regularization to improve generalization of the future frame prediction task. Experimental results on phonetic classification, speech recognition, and speech translation not only support the hypothesis, but also demonstrate the effectiveness of our approach in learning representations that contain richer phonetic content.