Lantian Li 2015-11-30
1. Go on my deep speaker embedding tasks:
1). knowledge transfer for i-vector -- working with Zhiyuan Zhang.
2). metric learning using linear transform.
Writing the technical report. -- Done !
2. DNN-ivector -- on the training process.
3. SUSR paper.
The baseline of GMM-UBM and i-vector have been done.
As shown in CVSS 425. The i-vecotr + PLDA is much better than GMM-UBM.
1. Go on the task 2 and 3.