Abstract: In this paper, we present an unsupervised learning framework to address the problem of detecting spoken keywords. Without any transcription information, a Gaussian Mixture Model is trained ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results