1 paper · 1 filter
Yael Segal, Tzeviya Sylvia Fuchs, Joseph Keshet
In this paper, we propose to apply object detection methods from the vision domain on the speech recognition domain, by treating audio fragments as objects. More specifically, we p…