Establishing degrees of closeness between audio recordings along different dimensions using large-scale cross-lingual models
This research focuses on understanding the information encoded in speech processing by using vector representations of speech from a pretrained model. The study proposes an unsupervised method using ABX tests…
Continue reading