# Filterbank Learning for Noise-Robust Small-Footprint Keyword Spotting

arXiv: https://arxiv.org/abs/2211.10565 · HTML: https://mlsanity.com/paper/2211.10565/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 18 references that are in the map.
- First posted on arXiv 19 Nov 2022.

## Builds on (most cited first, 9 of 9)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition](https://mlsanity.com/paper/1804.03209/) (2018)
- [Speaker Recognition from Raw Waveform with SincNet](https://mlsanity.com/paper/1808.00158/) (2018)
- [Deep Residual Learning for Small-Footprint Keyword Spotting](https://mlsanity.com/paper/1710.10361/) (2017)
- [Broadcasted Residual Learning for Efficient Keyword Spotting](https://mlsanity.com/paper/2106.04140/) (2021)
- [Deep Spoken Keyword Spotting: An Overview](https://mlsanity.com/paper/2111.10592/) (2021)
- [Small-Footprint Keyword Spotting on Raw Audio Data with Sinc-Convolutions](https://mlsanity.com/paper/1911.02086/) (2019)
- [Exploring Filterbank Learning for Keyword Spotting](https://mlsanity.com/paper/2006.00217/) (2020)
- [End-to-end Keyword Spotting using Neural Architecture Search and Quantization](https://mlsanity.com/paper/2104.06666/) (2021)

## Built on by (most cited first, 3 of 3)

- [Noise-Robust Keyword Spotting through Self-supervised Pretraining](https://mlsanity.com/paper/2403.18560/) (2024)
- [Robust Wake-Up Word Detection by Two-stage Multi-resolution Ensembles](https://mlsanity.com/paper/2310.11379/) (2023)
- [LearnAFE: Circuit-Algorithm Co-design Framework for Learnable Audio Analog Front-End](https://mlsanity.com/paper/2507.00755/) (2025)

## Abstract (arXiv)

In the context of keyword spotting (KWS), the replacement of handcrafted speech features by learnable features has not yielded superior KWS performance. In this study, we demonstrate that filterbank learning outperforms handcrafted speech features for KWS whenever the number of filterbank channels is severely decreased. Reducing the number of channels might yield certain KWS performance drop, but also a substantial energy consumption reduction, which is key when deploying common always-on KWS on low-resource devices. Experimental results on a noisy version of the Google Speech Commands Dataset show that filterbank learning adapts to noise characteristics to provide a higher degree of robustness to noise, especially when dropout is integrated. Thus, switching from typically used 40-channel log-Mel features to 8-channel learned features leads to a relative KWS accuracy loss of only 3.5% while simultaneously achieving a 6.3x energy consumption reduction.