Gated Embeddings in End-to-End Speech Recognition for Conversational-Context Fusion
arXiv:1906.11604
Abstract
We present a novel conversational-context aware end-to-end speech recognizer based on a gated neural network that incorporates conversational-context/word/speech embeddings. Unlike conventional speech recognition models, our model learns longer conversational-context information that spans across sentences and is consequently better at recognizing long conversations. Specifically, we propose to use the text-based external word and/or sentence embeddings (i.e., fastText, BERT) within an end-to-end framework, yielding a significant improvement in word error rate with better conversational-context representation. We evaluated the models on the Switchboard conversational speech corpus and show that our model outperforms standard end-to-end speech recognition models.
arXiv commentACL 2019
References in corpus (10)
- Sequence to Sequence Learning with Neural Networks
- ADADELTA: An Adaptive Learning Rate Method
- On the difficulty of training Recurrent Neural Networks
- Attention-Based Models for Speech Recognition
- Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- Bidirectional Attention Flow for Machine Comprehension
- Improved training of end-to-end attention models for speech recognition
- Advances in All-Neural Speech Recognition
- Larger-Context Language Modelling
- Direct Acoustics-to-Word Models for English Conversational Speech Recognition
Cited by in corpus (7)
- Adapting Long Context NLM for ASR Rescoring in Conversational Agents
- Cross-utterance Reranking Models with BERT and Graph Convolutional Networks for Conversational Speech Recognition
- Cross-Utterance Language Models with Acoustic Error Sampling
- Cross-Attention End-to-End ASR for Two-Party Conversations
- Distilling the Knowledge of BERT for Sequence-to-Sequence ASR
- LSTM-LM with Long-Term History for First-Pass Decoding in Conversational Speech Recognition
- Transformer Language Models with LSTM-based Cross-utterance Information Representation