# Normalise or condition? Noise-floor front-ends for on-board keyword spotting under UAV rotor ego-noise

arXiv: https://arxiv.org/abs/2609.31699 · HTML: https://mlsanity.com/paper/2609.31699/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 29 references that are in the map.
- First posted on arXiv 19 Sep 2026.

## Builds on (most cited first, 15 of 15)

- [FiLM: Visual Reasoning with a General Conditioning Layer](https://mlsanity.com/paper/1709.07871/) (2017)
- [SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition](https://mlsanity.com/paper/1904.08779/) (2019)
- [Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition](https://mlsanity.com/paper/1804.03209/) (2018)
- [Keyword Transformer: A Self-Attention Model for Keyword Spotting](https://mlsanity.com/paper/2104.00769/) (2021)
- [Broadcasted Residual Learning for Efficient Keyword Spotting](https://mlsanity.com/paper/2106.04140/) (2021)
- [Trainable Frontend For Robust and Far-Field Keyword Spotting](https://mlsanity.com/paper/1607.05666/) (2016)
- [Performance Evaluation of Deep Learning for Tree Branch Segmentation in Autonomous Forestry Systems](https://mlsanity.com/paper/2512.05418/) (2025)
- [YOLO and SGBM Integration for Autonomous Tree Branch Detection and Depth Estimation in Radiata Pine Pruning Applications](https://mlsanity.com/paper/2512.05412/) (2025)
- [Noise-Robust Keyword Spotting through Self-supervised Pretraining](https://mlsanity.com/paper/2403.18560/) (2024)
- [NTC-KWS: Noise-aware CTC for Robust Keyword Spotting](https://mlsanity.com/paper/2412.12614/) (2024)
- [Evaluating Voice Command Pipelines for Drone Control: From STT and LLM to Direct Classification and Siamese Networks](https://mlsanity.com/paper/2407.08658/) (2024)
- [AdaKWS: Towards Robust Keyword Spotting with Test-Time Adaptation](https://mlsanity.com/paper/2505.14600/) (2025)
- [DroneAudioset: An Audio Dataset for Drone-based Search and Rescue](https://mlsanity.com/paper/2510.15383/) (2025)
- [ImKWS: Test-Time Adaptation for Keyword Spotting with Class Imbalance](https://mlsanity.com/paper/2603.05821/) (2026)
- [End-to-End Voice Intent Recognition for Spontaneous Human-Drone Interaction with Naive Users](https://mlsanity.com/paper/2606.24910/) (2026)

## Abstract (arXiv)

A microphone on the airframe of a small multi-rotor UAV is dominated by rotor ego-noise, so spoken flight commands arrive at negative signal-to-noise ratio (SNR). We study small-footprint keyword spotting (KWS) for a ten-word command vocabulary under real ego-noise, training on one quadrotor and testing on another. Besides per-clip accuracy we measure the streaming false-alarm rate on 4.4 h of continuous rotor noise. We compare classical noise-robust front-ends (CMN, PCEN, spectral subtraction), test-time adaptation, and two front-ends that track the per-band ego-noise floor over the two seconds preceding the decision window and either subtract it (normalisation) or feed it to the network as a second input channel (conditioning). Per clip, all front-ends look alike: +2-3 points on average, up to +14 at -15 dB. On continuous rotor noise they differ sharply. Normalising front-ends fire about ten times more often than plain log-mel at the same threshold and end up below it at a budget of one false alarm per hour (-9 points at 0 dB). Conditioning keeps the baseline's false-alarm rate and turns its gain into detections (+8 points at -10 dB over three seeds); on top of PCEN it gives the best per-clip accuracy and false-alarm rate, and a level-anchored variant is also invariant to the microphone gain. Real drone+interferer recordings expose the remaining failure mode, environmental sounds and bystander speech, which training negatives halve. A single script reproduces all on-device numbers on the NVIDIA Jetson Orin NX flight computer, where the complete pipeline costs 3 ms per 100 ms hop on the CPU.