Toward Better Speaker Embeddings: Automated Collection of Speech Samples From Unknown Distinct Speakers | IEEE Conference Publication | IEEE Xplore