Wake Word Detection with Streaming Transformers
This work provides an incremental improvement in wake word detection performance for users of voice-activated devices.
This paper explores the application of streaming Transformers for wake word detection, addressing the non-streaming nature and quadratic complexity of vanilla Transformers. The proposed Transformer model achieved a 25% reduction in false rejection rate compared to a baseline convolutional network on the Mobvoi wake word dataset.
Modern wake word detection systems usually rely on neural networks for acoustic modeling. Transformers has recently shown superior performance over LSTM and convolutional networks in various sequence modeling tasks with their better temporal modeling power. However it is not clear whether this advantage still holds for short-range temporal modeling like wake word detection. Besides, the vanilla Transformer is not directly applicable to the task due to its non-streaming nature and the quadratic time and space complexity. In this paper we explore the performance of several variants of chunk-wise streaming Transformers tailored for wake word detection in a recently proposed LF-MMI system, including looking-ahead to the next chunk, gradient stopping, different positional embedding methods and adding same-layer dependency between chunks. Our experiments on the Mobvoi wake word dataset demonstrate that our proposed Transformer model outperforms the baseline convolution network by 25% on average in false rejection rate at the same false alarm rate with a comparable model size, while still maintaining linear complexity w.r.t. the sequence length.