“Sinovoice-2014-03-04”版本间的差异

来自cslt Wiki
跳转至： 导航、 搜索
 
（相同用户的2个中间修订版本未显示）
第19行： 第19行：
 
===470 + 300h + BJ mobile 105h training===
 
===470 + 300h + BJ mobile 105h training===
  
−
 
+
<pre>
 
Training condition                    NO NOISE        NOISE in LM      opt noise  NOISE LM + opt noise         
 
Training condition                    NO NOISE        NOISE in LM      opt noise  NOISE LM + opt noise         
  
 
No noise:                                30.61%          -                    -            -
 
No noise:                                30.61%          -                    -            -
 
noise phone added:                      31.88%          30.76%              31.27%        31.07
 
noise phone added:                      31.88%          30.76%              31.27%        31.07
−
 
+
</pre>
  
 
===BJ mobile incremental training===
 
===BJ mobile incremental training===
  
 
(1) Original 470 + 300 model: 30.24% WER
 
(1) Original 470 + 300 model: 30.24% WER
−
 
+
<pre>
 
MPE2      MPE3        MPE3+iLM      MPE4+iLM
 
MPE2      MPE3        MPE3+iLM      MPE4+iLM
 
27.01%    26.72%      25.09%        24.53%
 
27.01%    26.72%      25.09%        24.53%
 +
</pre>
  
  
第166行： 第167行：
 
|500ms        ||  28.29  ||  31.28  || 30.83 ||  30.22 ||  29.50  || 29.32  || 29.36
 
|500ms        ||  28.29  ||  31.28  || 30.83 ||  30.22 ||  29.50  || 29.32  || 29.36
 
|-
 
|-
−
}
+
|}
  
 
* CMN code delivery
 
* CMN code delivery
 
* Online model adaptation
 
* Online model adaptation

2014年3月4日 (二) 08:47的最后版本

DNN training

Environment setting

  • Schedule cluster poweroff on 3/06, construct RAID-0 on train212
  • The new RAID-0 used three new disks on train212
  • Change nfs names: disk1 -> /nfs/disk212, the raid disks: /nfs/raid212, disk2->/nfs/raid215

Corpora

  • PICC data are under labeling (200h), ready in one week.
  • 105h data from BJ mobile
  • 127h Hubei telecom
  • Now totally 1121h (470 + 346 + 105 + 200) telephone speech will be ready soon.
  • 16k 6000h data: 978h online data from DataTang + 656h online mobile data + 4300h recording data

Telephone model training

470 + 300h + BJ mobile 105h training

Training condition                    NO NOISE        NOISE in LM       opt noise   NOISE LM + opt noise        

No noise:                                30.61%           -                    -             -
noise phone added:                       31.88%          30.76%              31.27%         31.07

BJ mobile incremental training

(1) Original 470 + 300 model: 30.24% WER 

MPE2      MPE3         MPE3+iLM       MPE4+iLM
27.01%     26.72%       25.09%         24.53%


6000 hour 16k training

Training progress

  • Ran CE DNN to iteration 11 (8400 states, 80000 pdf)
  • Testing results go down to 12.49% WER (Sinovoice results: 10.49).
Model WER RT
small LM, it 4, -5/-9 15.80 1.18
large LM, it 4, -5/-9 15.30 1.50
large LM, it 4, -6/-9 15.36 1.30
large LM, it 4, -7/-9 15.25 1.30
large LM, it 5, -5/-9 14.17 1.10
large LM, it 5, -5/-10 13.77 1.29
large LM, it 6, -5/-9 13.64 1.12
large LM, it 6, -5/-10 13.25 1.33
large LM, it 7, -5/-9 13.29 1.12
large LM, it 7, -5/-10 12.87 1.17
large LM, it 8, -5/-9 13.09 -
large LM, it 8, -5/-10 12.69 -
large LM, it 9, -5/-9 12.87 -
large LM, it 9, -5/-10 12.55 -
large LM, it 10, -5/-9 12.83 1.51
large LM, it 10, -5/-10 12.48 1.65
large LM, it 11, -5/-9 12.87 1.61
large LM, it 11, -5/-10 12.46 1.28
large LM, it 12, -5/-9 12.91 1.61
large LM, it 12, -5/-10 12.49 1.28
  • xEnt training is done

Training Analysis

  • Shared tree GMM model training completed, WER% is similar to non-shared model .
  • Selected 100h online data, trained two systems: (1) di-syllable system (2) jt-phone system 
        di-syl      jt-ph
Xent    15.42%      14.78%       
MPE1    14.46%      14.23%
MPE2    14.22%      14.09%
MPE3    14.26%      13.80%
MPE4    14.24%      13.68%
  • Xiaoming is working on HTK training on the same database

Hybrid training

  • Receipe
  • 100h MPE training
  • 1700h MPE alignment/lattice
  • 1700h MPE training
  • 1 week to complete 3 MPE iterations
  • MPE2 result: 1e-9: 10.67% (8.61%), 1e-10: 10.34% (8.27%)


Auto Transcription

PICC

  • PICC development set decoding obtained 45% WER.
  • PICC auto-trans incremental DT training completed 
Threshold  WER
org:     45.03%
0.9:     41.89%
0.8:     41.64%
  • Current training data with 0.8 involve 80k sentences, amounting to about 60h data.
  • Sampling 60h labelled data to enrich the training
  • Prepare to compare the unsupervised incremental training and supervised training

Hubei telecom

  • Hubei telecom data (127 h), retrieve 60k sentence by conf thred=0.9, amounting to 50% 
xEnt org:  -             wer_15  29.05
MPE iter1：wer_14 29.23；wer_15 29.38
MPE iter2：wer_14 29.05；wer_15 29.11
MPE iter3：wer_14 29.32；wer_15 29.28
MPE iter4：wer_14 29.29；wer_15 29.28
  • retrieve 30k sentences by conf thred=0.95, amounting to 25%, plus the original 770h data 
xEnt org:     -             wer_15  29.05
MPE iter1:    -             wer_15: 29.36


DNN Decoder

Online decoder

  • Various CMN implementation test
  • 200ms/500ms frame block adaptation
  • 10ms frame block adaptation: totally wrong
prior weight -1 1 5 10 20 50 100
200ms 28.29 37.53 35.50 34.08 32.90 32.30 32.77
500ms 28.29 31.28 30.83 30.22 29.50 29.32 29.36
  • CMN code delivery
  • Online model adaptation
取自“http://cslt.org/mediawiki/index.php?title=Sinovoice-2014-03-04&oldid=9256”