Zero-Shot KWS for Children's Speech using Layer-Wise Features from SSL Models

Open in new window