Yeonsu Kim, Solee Im, Jihyun Lee, Wonjun Lee · ACM Symposium on User Interface Software and Technology (UIST) 2026 · 2026
DOI: 10.1145/3830397.3841824
Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).
We present Wake2Adapt, a system that personalizes voice assistants using wake-word utterances users already produce. Using recognition of Korean proper nouns uttered by non-native speakers as a representative task, we compare four wake-word types. Korean wake words improved recognition, whereas existing English wake words did not. Ablations show that the gains stem from phoneme coverage rather than length, and that a single utterance suffices.
arXiv (Cornell University) 2026 · 0 citations
No comments yet — start the discussion below.