Filter-based Discriminative Autoencoders for Children Speech Recognition
This work addresses the challenge of diverse children's speech for speech recognition systems, representing an incremental improvement with specific gains.
The paper tackles children speech recognition by proposing a filter-based discriminative autoencoder to handle speaker and pitch variability, achieving a 7.8% relative WER reduction on the CMU Kids corpus and improved performance in domain adaptation tasks.
Children speech recognition is indispensable but challenging due to the diversity of children's speech. In this paper, we propose a filter-based discriminative autoencoder for acoustic modeling. To filter out the influence of various speaker types and pitches, auxiliary information of the speaker and pitch features is input into the encoder together with the acoustic features to generate phonetic embeddings. In the training phase, the decoder uses the auxiliary information and the phonetic embedding extracted by the encoder to reconstruct the input acoustic features. The autoencoder is trained by simultaneously minimizing the ASR loss and feature reconstruction error. The framework can make the phonetic embedding purer, resulting in more accurate senone (triphone-state) scores. Evaluated on the test set of the CMU Kids corpus, our system achieves a 7.8% relative WER reduction compared to the baseline system. In the domain adaptation experiment, our system also outperforms the baseline system on the British-accent PF-STAR task.