End2End Acoustic to Semantic Transduction
arXiv:2102.01013 · doi:10.1109/ICASSP39728.2021.9413581
Abstract
In this paper, we propose a novel end-to-end sequence-to-sequence spoken language understanding model using an attention mechanism. It reliably selects contextual acoustic features in order to hypothesize semantic contents. An initial architecture capable of extracting all pronounced words and concepts from acoustic spans is designed and tested. With a shallow fusion language model, this system reaches a 13.6 concept error rate (CER) and an 18.5 concept value error rate (CVER) on the French MEDIA corpus, achieving an absolute 2.8 points reduction compared to the state-of-the-art. Then, an original model is proposed for hypothesizing concepts and their values. This transduction reaches a 15.4 CER and a 21.6 CVER without any new type of context.
arXiv commentAccepted at IEEE ICASSP 2021
References in corpus (7)
- Neural Machine Translation by Jointly Learning to Align and Translate
- On Using Monolingual Corpora in Neural Machine Translation
- End-to-end Continuous Speech Recognition using Attention-based Recurrent NN: First Results
- vq-wav2vec: Self-Supervised Learning of Discrete Speech Representations
- Unsupervised speech representation learning using WaveNet autoencoders
- Dialogue history integration into end-to-end signal-to-concept spoken language understanding systems
- CSTNet: Contrastive Speech Translation Network for Self-Supervised Speech Representation Learning