Recommended Citation
Researchers using this dataset in their publications are requested to cite the following paper:
Z.-W. Tan, V. G. Reju, R. C. Tewari, R. Ding, P. A. Naylor, and A. W. H. Khong, "ICASSP 2026 Radar Acoustic Speech Enhancement (RASE) Challenge," in Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2026.
DOI: https://doi.org/10.1109/ICASSP55912.2026.11463564
Official Challenge Website
For information about the ICASSP 2026 RASE Challenge, including the challenge overview, evaluation protocols, Github repo with baseline, participation guidelines, and related resources, please visit:
https://rase-challenge.github.io/RASE2026-Challenge/
The ICASSP 2026 Radar Acoustic Speech Enhancement (RASE) Challenge Dataset is a benchmark resource developed to advance research in radar-based speech sensing and enhancement.
The challenge focuses on recovering intelligible speech from radar measurements of acoustic vibrations, including vibrations captured through glass barriers. The dataset supports the development, evaluation, and comparison of machine learning and signal processing approaches for radar-acoustic speech enhancement.
Dataset Release:
This release contains the training and evaluation datasets used in the ICASSP 2026 RASE Challenge. The test dataset is not included and will be released separately at a later date.
This dataset accompanies the official RASE Challenge paper published at the 2026 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2026).