MALEFA: Multi-grAnularity Learning and Effective False Alarm Suppression for Zero-Shot Keyword Spotting
Lo-Ya Li, Tien-Hong Lo, Jeih-Weih Hung, Shih-Chieh Huang, Berlin Chen · 2026
Read MALEFA: Multi-grAnularity Learning and Effective False Alarm Suppression for Zero-Shot Keyword Spotting by Lo-Ya Li, Tien-Hong Lo, Jeih-Weih Hung, Shih-Chieh Huang, Berlin Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-03689.txt.
Description
Lo-Ya Li and colleagues built a light system that spots user-chosen spoken keywords without training data for each new word. MALEFA learns matches at both whole-utterance and sound-unit levels, cutting false alarms on similar-sounding words. It reached 90 percent accuracy on four public sets, with a false-alarm rate of 0.007 percent on AMI, and runs fast enough for small devices.
Catalog record
| CREATOR | Lo-Ya Li, Tien-Hong Lo, Jeih-Weih Hung, Shih-Chieh Huang, Berlin Chen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17602 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.03689