“2014-10-20”版本间的差异
来自cslt Wiki
| 第61行： | 第61行： | ||
* Add more silence tag "#" in pure-silence utterance text(train). | * Add more silence tag "#" in pure-silence utterance text(train). | ||
:* xEntropy model be training | :* xEntropy model be training | ||
| − | : need to test baseline. | + | :* need to test baseline. |
* Sum all sil-pdf as the silence posterior probability. | * Sum all sil-pdf as the silence posterior probability. | ||
| 第75行： | 第75行： | ||
:* Various the used Chinese trained DNN layer numbers. | :* Various the used Chinese trained DNN layer numbers. | ||
:** feature_transform = 6000h_transform + 6000_N*hidden-layers | :** feature_transform = 6000h_transform + 6000_N*hidden-layers | ||
| − | nnet.init = | + | nnet.init = random (4-N)*hidden-layers + output-layer |
| − | + | ||
| − | + | ||
| N / learn_rate | 0.008 | 0.001 | 0.0001 | | | N / learn_rate | 0.008 | 0.001 | 0.0001 | | ||
| baseline | 17.00(14*2h) | | | | | baseline | 17.00(14*2h) | | | | ||
| 第88行： | 第86行： | ||
nnet.init = random (4-N)*hidden-layers + output-layer | nnet.init = random (4-N)*hidden-layers + output-layer | ||
Note: This is reproduced Yinshi's experiment | Note: This is reproduced Yinshi's experiment | ||
| − | |||
| N / learn_rate | 0.008 | 0.001 | 0.0001 | | | N / learn_rate | 0.008 | 0.001 | 0.0001 | | ||
| baseline | 17.00 | | | | | baseline | 17.00 | | | | ||
2014年10月20日 (一) 07:06的版本
Speech Processing
AM development
Contour
- NAN problem
- nan recurrence
------------------------------------------------------------ grid/atr. | Reproducible | add. ------------------------------------------------------------ grid-10 | yes | ------------------------------------------------------------ grid-12 | no | "nan" in different position ------------------------------------------------------------ grid-14 | yes | ------------------------------------------------------------
Sparse DNN
- Performance improvement found when pruned slightly
- Experiments show that
- Suggest to use TIMIT / AURORA 4 for training
RNN AM
- Initial test on WSJ , leads to out-memory.
- Using AURORA 4 short-sentence with a smaller number of targets.
Noise training
- First draft of the noisy training journal paper
- Paper Correction (Yinshi, Liuchao, Lin Yiye), be going.
Drop out & Rectification & convolutive network
- Drop out
- dataset:wsj, testset:eval92
std | dropout0.4 | dropout0.5 | dropout0.7 | dropout0.8
-------------------------------------------------------------
4.5 | 5.39 | 4.80 | 4.36 | -
- Test on noisy AURORA4 dataset
std | dropout0.4 | dropout0.5 | dropout0.7 | dropout0.8
-------------------------------------------------------------
6.05 | - | - | - | -
- Continue the droptout on normal trained XEnt NNET , eg wsj. (+)
- Draft the dropout-DNN weight distribution. (+)
- Rectification
- Still NAN error, need to debug. (+)
- MaxOut (+)
- Convolutive network
- Test more configurations
- Yiye will work on CNN
- Reading CNN tutorial
Denoising & Farfield ASR
- ICASSP paper submitted.
VAD
- Add more silence tag "#" in pure-silence utterance text(train).
- xEntropy model be training
- need to test baseline.
- Sum all sil-pdf as the silence posterior probability.
- Program done, to tune the threshold
Speech rate training
- Seems ROS model is superior to the normal one with faster speech
- Suggest to extract speech data of different ROS, construct a new test set(+)
- Suggest to use Tencent training data(+)
low resource language AM training
- Use Chinese NN as initial NN, change the last layer
- Various the used Chinese trained DNN layer numbers.
- feature_transform = 6000h_transform + 6000_N*hidden-layers
- Various the used Chinese trained DNN layer numbers.
nnet.init = random (4-N)*hidden-layers + output-layer | N / learn_rate | 0.008 | 0.001 | 0.0001 | | baseline | 17.00(14*2h) | | | | 4 | 17.75(9*0.6h) | 18.64 | | | 3 | 16.85 | | | | 2 | 16.69 | | | | 1 | 16.87 | | | | 0 | 16.88 | | |
- feature_transform = uyghur_transform + 6000_N*hidden-layers
nnet.init = random (4-N)*hidden-layers + output-layer Note: This is reproduced Yinshi's experiment | N / learn_rate | 0.008 | 0.001 | 0.0001 | | baseline | 17.00 | | | | 4 | 28.23 | 30.72 | 37.32 | | 3 | 22.40 | | | | 2 | 19.76 | | | | 1 | 17.41 | | | | 0 | | | |
- feature_transform = 6000_transform + 6000_N*hidden-layers
nnet.init = uyghur (4-N)*hidden-layers + output-layer | N / learn_rate | 0.008 | 0.001 | 0.0001 | | baseline | 17.00 | | | | 4 | 17.80 | 18.55 | 21.06 | | 3 | 16.89 | 17.64 | | | 2 | | | | | 1 | | | | | 0 | | | |
Scoring
- global scoring done.
- Pitch & rhythm done, need testing
- Harmonics program done, experiment to be done.
Confidence
- Reproduce the experiments on fisher dataset.
- Use the fisher DNN model to decode all-wsj dataset
Speaker ID
- Preparing GMM-based server.
Emotion detection
- Sinovoice is implementing the server
Text Processing
LM development
Domain specific LM
h2. ngram generation is on going h2. look the memory and baidu_hi done
h2. NUM tag LM:
- maxi work is released.
- yuanbin continue the tag lm work.
- add the ner to tag lm .
- Boost specific words like wifi if TAG model does not work for a particular word.
Word2Vector
W2V based doc classification
- Initial results variable Bayesian GMM obtained. Performance is not as good as the conventional GMM.
- Non-linear inter-language transform: English-Spanish-Czch: wv model training done, transform model on investigation
- SSA-based local linear mapping still on running.
- k-means classes change to 2.
- Knowledge vector started
- format the data
- Character to word conversion
- prepare the task: word similarity
- prepare the dict.
- Google word vector train
- improve the sampling method
RNN LM
- rnn
- lstm+rnn
- install the tool and prepare the data of wsj
- prepare the baseline.
Translation
- v3.0 demo released
- still slow
- re-segment the word using new dictionary.
- check new data.
QA
- search method:
- add the vsm and BM25 to improve the search. and the strategy of selecting the answer
- segment the word using minimum granularity for lucene index and bag-of-words method.
- new inter will install SEMPRE
