Streaming Audio-Visual Speech Recognition with Alignment Regularization
This work improves speech recognition accuracy in noisy environments for applications like real-time communication, though it is incremental as it builds on existing streaming and AV-ASR methods.
The paper tackled the problem of streaming audio-visual speech recognition by proposing a hybrid CTC/attention neural network with alignment regularization to synchronize audio and visual encoders, achieving state-of-the-art word error rates of 2.0% (offline) and 2.6% (online) on the LRS3 dataset without external training data.
In this work, we propose a streaming AV-ASR system based on a hybrid connectionist temporal classification (CTC)/attention neural network architecture. The audio and the visual encoder neural networks are both based on the conformer architecture, which is made streamable using chunk-wise self-attention (CSA) and causal convolution. Streaming recognition with a decoder neural network is realized by using the triggered attention technique, which performs time-synchronous decoding with joint CTC/attention scoring. Additionally, we propose a novel alignment regularization technique that promotes synchronization of the audio and visual encoder, which in turn results in better word error rates (WERs) at all SNR levels for streaming and offline AV-ASR models. The proposed AV-ASR model achieves WERs of 2.0% and 2.6% on the Lip Reading Sentences 3 (LRS3) dataset in an offline and online setup, respectively, which both present state-of-the-art results when no external training data are used.