Dual-Attention Neural Transducers for Efficient Wake Word Spotting in Speech Recognition
Not on the map. Why?
Abstract (arXiv)
We present dual-attention neural biasing, an architecture designed to boost Wake Words (WW) recognition and improve inference time latency on speech recognition tasks. This architecture enables a dynamic switch for its runtime compute paths by exploiting WW spotting to select which branch of its attention networks to execute for an input audio frame. With this approach, we effectively improve WW spotting accuracy while saving runtime compute cost as defined by floating point operations (FLOPs). Using an in-house de-identified dataset, we demonstrate that the proposed dual-attention network can reduce the compute cost by 90% for WW audio frames, with only 1% increase in the number of parameters. This architecture improves WW F1 score by 16% relative and improves generic rare word error rate by 3% relative compared to the baselines.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 30 of its 36 references that are in the map. First posted on arXiv 3 Apr 2023.
Builds on
30 of 36 references.
References · 30
- 2017Attention Is All You Needcited by 59,932
- 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,898
- 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,838
- 2020Rethinking Attention with Performerscited by 1,584
- 2020Linformer: Self-Attention with Linear Complexitycited by 1,462
- 2018Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidatescited by 632
- 2017To prune, or not to prune: exploring the efficacy of pruning for model compressioncited by 596
- 2015Listen, Attend and Spellcited by 542
- 2020Efficient Content-Based Sparse Attention with Routing Transformerscited by 469
- 2016Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activationscited by 413
- 2015LSTM: A Search Space Odysseycited by 315
- 2020Transformer Transducer: A Streamable Speech Recognition Model with Transformer Encoders and RNN-T Losscited by 232
18 more
- 2018Streaming End-to-end Speech Recognition For Mobile Devicescited by 230
- 2018Deep context: end-to-end contextual speech recognitioncited by 136
- 2019Transformer-Transducer: End-to-End Speech Recognition with Self-Attentioncited by 68
- 2020A Better and Faster End-to-End Model for Streaming ASRcited by 68
- 2017Simple Recurrent Units for Highly Parallelizable Recurrencecited by 60
- 2022Contextual Adapters for Personalized Speech Recognition in Neural Transducerscited by 57
- 2021Context-Aware Transformer Transducer for Speech Recognitioncited by 56
- 2017A Regularized Framework for Sparse and Structured Neural Attentioncited by 38
- 2016Neural Speech Recognizer: Acoustic-to-Word LSTM Model for Large Vocabulary Speech Recognitioncited by 33
- 2021Personalization Strategies for End-to-End Speech Recognition Systemscited by 26
- 2019Self-Attention Transducers for End-to-End Speech Recognitioncited by 24
- 2017No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Modelscited by 13
- 2021Dynamic Encoder Transducer: A Flexible Solution For Trading Off Accuracy For Latencycited by 9
- 2021Amortized Neural Networks for Low-Latency Speech Recognitioncited by 9
- 2022ConvRNN-T: Convolutional Augmented Recurrent Neural Network Transducers for Streaming Speech Recognitioncited by 8
- 2021Bifocal Neural ASR: Exploiting Keyword Spotting for Inference Optimizationcited by 7
- 2021Exploiting Large-scale Teacher-Student Training for On-device Acoustic Modelscited by 5
- 2022Compute Cost Amortized Transformer for Streaming ASRcited by 2
Built on by
Citing papers per year: 2024: 2.
2 papers.