140701 - Bie Fanhu

来自cslt Wiki
2014年7月1日 (二) 04:25Biefh讨论 | 贡献的版本

(差异) ←上一版本 | 最后版本 (差异) | 下一版本→ (差异)
跳转至: 导航搜索

Last Week:

1. Increase the data for training the DNN net work to reconstruct the truncated speech data.

2. Complete the DNN reconstruct work to other truncated rate.

3. Code for the DNN based ivector extraction.

This Week:

1. Complete DNN based ivector extraction.

2. Complete the different rate truncated data reconstruct.