Max-Pooling Loss Training of Long Short-Term Memory Networks for Small-Footprint Keyword Spotting
This addresses small-footprint keyword spotting for embedded devices, but it is incremental as it builds on existing LSTM and loss function techniques.
The paper tackles keyword spotting with low resource requirements by proposing a max-pooling loss function for training LSTM networks, resulting in a 67.6% relative reduction in AUC compared to a baseline feed-forward DNN.
We propose a max-pooling based loss function for training Long Short-Term Memory (LSTM) networks for small-footprint keyword spotting (KWS), with low CPU, memory, and latency requirements. The max-pooling loss training can be further guided by initializing with a cross-entropy loss trained network. A posterior smoothing based evaluation approach is employed to measure keyword spotting performance. Our experimental results show that LSTM models trained using cross-entropy loss or max-pooling loss outperform a cross-entropy loss trained baseline feed-forward Deep Neural Network (DNN). In addition, max-pooling loss trained LSTM with randomly initialized network performs better compared to cross-entropy loss trained LSTM. Finally, the max-pooling loss trained LSTM initialized with a cross-entropy pre-trained network shows the best performance, which yields $67.6\%$ relative reduction compared to baseline feed-forward DNN in Area Under the Curve (AUC) measure.