Improving Speaker Identification for Shared Devices by Adapting Embeddings to Speaker Subsets

Re-assign community