Cslt：/* Optimal phoneset */

2014-01-10T13:38:58Z

‎Optimal phoneset

Cslt：/* Speech QA */

2014-01-10T01:30:43Z

‎Speech QA

Cslt：以内容“== AM development == === Sparse DNN === * Optimal Brain Damage(OBD). # Online OBD held. # OBD + L1 norm start to investigation. * Efficient computing # Conducti...”创建新页面

2014-01-10T01:26:14Z

以内容“== AM development == === Sparse DNN === * Optimal Brain Damage(OBD). # Online OBD held. # OBD + L1 norm start to investigation. * Efficient computing # Conducti...”创建新页面

新页面

== AM development ==

=== Sparse DNN ===

* Optimal Brain Damage(OBD).

# Online OBD held.
# OBD + L1 norm start to investigation.

* Efficient computing

# Conducting rearrangement the matrix structure and compose zero blocks by some smart approaches, leading to better computing speed.

=== Efficient DNN training ===

# Asymmetric window: Great improvement on training set(WER 34% to 24%), however the improvement is lost on test. Overfitting?
# Fbank feature used to train GMM+DNN, leads to very high training Acc, but reduces accuracy on test.
# Fbank energy trim. With threshold=10, we can get performance gain in a safe way. Larger thresholds may lead to performance reduction.[http://192.168.0.50:3000/series/?action=view&series=201,201.0,201.1,201.2,201.3|click here]

=== Optimal phoneset===

* Ch/En training with concatenated phone set is completed.
* Errors in testing.

===Engine optimization===

* Investigating LOUDS FST. On progress.

==LM development==

===NN LM===

* Working on NN LM based on word2vector.
* Reading more materials on word2vector.

==Embedded development==

* Liuchao's cellphone, Qualcomm Snapdragon Krait MSM8960 @ 1.5GHz, using 1 core
small nnet 100/600/600/600/600/1264 with MFCC input
* Work on layer-by-layer DNN training

==Speech QA==

* Use N-best to expand match in QA. Better performance were obtained.
:* 1-best matches 96/121
:* 10-best matches 102/121

* Use N-best to recover errors in entity check. Working on.
* Use Pinyin to recover errors in entity check. Future work.

←上一版本		2014年1月10日 (五) 01:30的版本
第50行：		第50行：
	* Use N-best to recover errors in entity check. Working on.		* Use N-best to recover errors in entity check. Working on.
	* Use Pinyin to recover errors in entity check. Future work.		* Use Pinyin to recover errors in entity check. Future work.
		+	* Investigate some errors in entity-based LM.

@@ 第22行： / 第22行： @@
 * Ch/En training with concatenated phone set is completed.
-* Errors in testing.
+* Checking errors in testing.
 ===Engine optimization===

2014-01-10 - 版本历史

Cslt：/* Optimal phoneset */

Cslt：/* Speech QA */

Cslt：以内容“== AM development == === Sparse DNN === * Optimal Brain Damage(OBD). # Online OBD held. # OBD + L1 norm start to investigation. * Efficient computing # Conducti...”创建新页面