Wav2Seq: Pre-training Speech-to-Text Encoder-Decoder Models Using Pseudo Languages
This addresses the challenge of pre-training both encoder and decoder components in speech-to-text models, offering a low-cost solution that benefits tasks like ASR, spoken named entity recognition, and translation, though it is incremental as it builds on existing encoder-decoder frameworks.
The paper tackles the problem of pre-training encoder-decoder models for speech data by introducing Wav2Seq, a self-supervised method that uses pseudo languages to transcribe audio into pseudo subword sequences, resulting in new state-of-the-art results for end-to-end spoken named entity recognition and consistent improvements on 20 language pairs for speech-to-text translation.
We introduce Wav2Seq, the first self-supervised approach to pre-train both parts of encoder-decoder models for speech data. We induce a pseudo language as a compact discrete representation, and formulate a self-supervised pseudo speech recognition task -- transcribing audio inputs into pseudo subword sequences. This process stands on its own, or can be applied as low-cost second-stage pre-training. We experiment with automatic speech recognition (ASR), spoken named entity recognition, and speech-to-text translation. We set new state-of-the-art results for end-to-end spoken named entity recognition, and show consistent improvements on 20 language pairs for speech-to-text translation, even when competing methods use additional text data for training. Finally, on ASR, our approach enables encoder-decoder methods to benefit from pre-training for all parts of the network, and shows comparable performance to highly optimized recent methods.