News-2018-01-30

来自cslt Wiki
2018年1月29日 (一) 23:13Cslt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

Lantian Li, Dong Wang, Yixiang Chen, Ying Shi, Zhiyuan Tang, "DEEP FACTORIZATION FOR SPEECH SIGNAL"

This paper describes how speech signals can be factorized into varios informative factors in the latent space.

More information about this research can be found in the project page

Lantian Li, Zhiyuan Tang, Dong Wang, "FULL-INFO TRAINING FOR DEEP SPEAKER FEATURE LEARNING"

It presents a new idea to learn 'pure features', by pushing all the discriminative information to feature learning. It provides stonger features than regular feature + softmax regression architecture.

More information about this research can be found in the project page


Miao Zhang, Xiaofei Kang, Yanqing Wang, Lantian Li, Zhiyuan Tang, Haisheng Dai, Dong Wang, "HUMAN AND MACHINE SPEAKER RECOGNITION BASED ON SHORT TRIVIAL EVENTS"

It presents how machines can discriminate speakers by trivial events such as laugh, cough, en.

More information about this research can be found in the project page.