Audio-Visual Decision Fusion for WFST-based and seq2seq Models
arXiv:2001.10832
Abstract
Under noisy conditions, speech recognition systems suffer from high Word Error Rates (WER). In such cases, information from the visual modality comprising the speaker lip movements can help improve the performance. In this work, we propose novel methods to fuse information from audio and visual modalities at inference time. This enables us to train the acoustic and visual models independently. First, we train separate RNN-HMM based acoustic and visual models. A common WFST generated by taking a special union of the HMM components is used for decoding using a modified Viterbi algorithm. Second, we train separate seq2seq acoustic and visual models. The decoding step is performed simultaneously for both modalities using shallow fusion while maintaining a common hypothesis beam. We also present results for a novel seq2seq fusion without the weighing parameter. We present results at varying SNR and show that our methods give significant improvements over acoustic-only WER.
Submitted for review to ICASSP 2020 on October 21st, 2019
References in corpus (7)
- Adam: A Method for Stochastic Optimization
- Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- Deep Audio-Visual Speech Recognition
- On Using Monolingual Corpora in Neural Machine Translation
- Improved training of end-to-end attention models for speech recognition
- LipNet: End-to-End Sentence-level Lipreading
- Towards better decoding and language model integration in sequence to sequence models