HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Conference papers

Keyword Spotting System using Low-complexity Feature Extraction and Quantized LSTM

Abstract : Long Short-Term Memory (LSTM) neural networks offer state-of-the-art results to compute sequential data and address applications like keyword spotting. Mel Frequency Cepstral Coefficients (MFCC) are the most common features used to train this neural network model. However, the complexity of MFCC coupled with highly optimized machine learning neural networks usually makes the MFCC feature extraction the most power-consuming block of the system. This paper presents a low complexity feature extraction method using a filter bank composed of 16 channels with a quality factor of 1.3 to compute a spectrogram. It shows that we can achieve an 89.45% accuracy on 12 classes of the Google Speech Command Dataset using an LSTM network of 64 hidden units with weights and activation quantized to 9 bits and inputs quantized to 8 bits.
Document type :
Conference papers
Complete list of metadata

Contributor : Antoine Frappé Connect in order to contact the contributor
Submitted on : Thursday, January 6, 2022 - 7:21:00 PM
Last modification on : Friday, May 20, 2022 - 9:23:40 AM
Long-term archiving on: : Thursday, April 7, 2022 - 8:05:57 PM


ICECS_HK_PostReview V2.pdf
Files produced by the author(s)




Kévin Hérissé, Benoit Larras, Antoine Frappé, Andreas Kaiser. Keyword Spotting System using Low-complexity Feature Extraction and Quantized LSTM. 28th IEEE International Conference on Electronics Circuits and Systems, ICECS 2021, Nov 2021, Dubai, United Arab Emirates. ⟨10.1109/ICECS53924.2021.9665486⟩. ⟨hal-03515704⟩



Record views


Files downloads