Unsupervised domain adaptation for speech recognition with unsupervised error correction
This addresses the domain adaptation problem for speech recognition systems, enabling better performance in new environments without requiring labeled data, though it is incremental as it builds on existing correction and adaptation techniques.
The paper tackles the problem of automatic speech recognition (ASR) performance degradation in unseen domains by proposing an unsupervised error correction method that uses only unlabeled target domain data, achieving a significant word error rate reduction and an additional 10% relative improvement when combined with other adaptation approaches.
The transcription quality of automatic speech recognition (ASR) systems degrades significantly when transcribing audios coming from unseen domains. We propose an unsupervised error correction method for unsupervised ASR domain adaption, aiming to recover transcription errors caused by domain mismatch. Unlike existing correction methods that rely on transcribed audios for training, our approach requires only unlabeled data of the target domains in which a pseudo-labeling technique is applied to generate correction training samples. To reduce over-fitting to the pseudo data, we also propose an encoder-decoder correction model that can take into account additional information such as dialogue context and acoustic features. Experiment results show that our method obtains a significant word error rate (WER) reduction over non-adapted ASR systems. The correction model can also be applied on top of other adaptation approaches to bring an additional improvement of 10% relatively.