search

modelscope / 3D-Speaker

A Repository for Single- and Multi-modal Speaker Verification, Speaker Recognition and Speaker Diarization
Apache License 2.0
1.02k stars 89 forks source link

SELF-DISTILLATION NETWORK WITH ENSEMBLE PROTOTYPES: LEARNING ROBUST SPEAKER REPRESENTATIONS WITHOUT SUPERVISION #115

Closed gancx closed 2 months ago

gancx commented 2 months ago

Thanks for your contribution to speaker verification and open-source code. I wonder if the bin/train_sdpn.py corresponds to the paper (SELF-DISTILLATION NETWORK WITH ENSEMBLE PROTOTYPES: LEARNING ROBUST SPEAKER REPRESENTATIONS WITHOUT SUPERVISION). They look quite similar but slightly different.

yfchenlucky commented 2 months ago

The code and the paper content do not match exactly but are very similar. The latest version of the paper will be open-sourced in June. Thank you for your attention. If it is helpful, please star and cite it.

gancx commented 2 months ago

The code and the paper content do not match exactly but are very similar. The latest version of the paper will be open-sourced in June. Thank you for your attention. If it is helpful, please star and cite it.

Thanks for your prompt reply.