Difference between revisions of "Lantian Li 2016-03-21"

From cslt Wiki
Jump to: navigation, search
(以“* Weekly Report :* Binary speaker embedding. Only need 'variable-sized block training'. :* Replay detection based on low-energy speech segments. :* Joint-training...”为内容创建页面)
 
 
Line 1: Line 1:
 
* Weekly Report
 
* Weekly Report
  
:* Binary speaker embedding. Only need 'variable-sized block training'.
+
:* Binary speaker embedding. Results is inconsistent.
  
 
:* Replay detection based on low-energy speech segments.
 
:* Replay detection based on low-energy speech segments.

Latest revision as of 00:57, 21 March 2016

  • Weekly Report
  • Binary speaker embedding. Results is inconsistent.
  • Replay detection based on low-energy speech segments.
  • Joint-training.here. Hold
  • Pair-wised speaker vector training.here Hold
  • Deep speaker embedding. Hold
  • This Week
  • Complete Interspeech papers.
  • Joint-training (speaker recognition)