Speaker Embedding for verification / identification task using Tensorflow
Index | filename | speakers | speaker_id | duration |
---|---|---|---|---|
0 | /home/manish_kumar/speaker_recognition/data/testutt1.wav | Jhon | 253 | 11.46 |
1 | /home/manish_kumar/speaker_recognition/data/testutt2.wav | Jack | 69 | 14.32 |
2 | /home/manish_kumar/speaker_recognition/data/testutt3.wav | Nat | 750 | 4.62 |
3 | /home/manish_kumar/speaker_recognition/data/testutt4.wav | David | 723 | 10.54 |
python train_softmax.py --chkpt chkpt1 --E_train 50 --lr 0.0001 --bs 64 --csv ted_train.csv