“ASR:2015-10-19”版本间的差异
来自cslt Wiki
（以“==Speech Processing == === AM development === ==== Environment ==== * repair laptop ==== RNN AM==== *train monophone RNN --zhiyuan :* decode using 5-gram :* the t...”为内容创建页面） |
（→Speech Processing） |
||
| 第3行： | 第3行： | ||
==== Environment ==== | ==== Environment ==== | ||
| − | |||
| − | |||
==== RNN AM==== | ==== RNN AM==== | ||
*train monophone RNN --zhiyuan | *train monophone RNN --zhiyuan | ||
| − | :* | + | :* end to end MPE |
| − | + | ||
| − | + | ||
* train RNN MPE using large dataset--mengyuan | * train RNN MPE using large dataset--mengyuan | ||
| − | + | :* try adaptation method using Daohangquan dataset | |
| − | :* try adaptation method | + | |
| − | + | ||
====Learning rate tunning==== | ====Learning rate tunning==== | ||
| 第22行： | 第16行： | ||
* hold | * hold | ||
* compute EER with kaldi | * compute EER with kaldi | ||
| + | |||
| + | ===Decision tree=== | ||
| + | * decision tree balance using 100h Chinese and 20h English --zhiyong | ||
====Data selection unsupervised learning==== | ====Data selection unsupervised learning==== | ||
| 第30行： | 第27行： | ||
====RNN-DAE(Deep based Auto-Encode-RNN)==== | ====RNN-DAE(Deep based Auto-Encode-RNN)==== | ||
| + | * hold | ||
* RNN-DAE has worse performance than DNN-DAE because training dataset is small | * RNN-DAE has worse performance than DNN-DAE because training dataset is small | ||
* extract real room impulse to generate WSJ reverberation data, and then train RNN-DAE | * extract real room impulse to generate WSJ reverberation data, and then train RNN-DAE | ||
===Ivector&Dvector based ASR=== | ===Ivector&Dvector based ASR=== | ||
| − | * | + | * learning from ivector --Lantian |
| − | :* | + | :* CNN ivector learning |
| + | :* DNN ivector learning | ||
* binary ivector | * binary ivector | ||
| − | * metric learning | + | * metric learning |
| + | |||
===language vector=== | ===language vector=== | ||
| − | |||
* write a paper--zhiyuan | * write a paper--zhiyuan | ||
| + | :*hold | ||
* language vector is added to multi hidden layers--zhiyuan | * language vector is added to multi hidden layers--zhiyuan | ||
* RNN language vector | * RNN language vector | ||
| 第49行： | 第49行： | ||
* multi-stream training --Sheng Su | * multi-stream training --Sheng Su | ||
:*two GPUs work well, but four GPUs divergent | :*two GPUs work well, but four GPUs divergent | ||
| − | |||
* kaldi-nnet3 --Xuewei | * kaldi-nnet3 --Xuewei | ||
===multi-task=== | ===multi-task=== | ||
| − | * | + | * test according to selt-information neural structure learning --mengyuan |
* speech rate learning --xiangyu | * speech rate learning --xiangyu | ||
2015年10月19日 (一) 08:42的版本
目录
- 1 Speech Processing
- 2 =Neutral picture style transfer
- 3 Text Processing
- 4 financial group
Speech Processing
AM development
Environment
RNN AM
- train monophone RNN --zhiyuan
- end to end MPE
- train RNN MPE using large dataset--mengyuan
- try adaptation method using Daohangquan dataset
Learning rate tunning
- sequence training -Xiangyu
Mic-Array
- hold
- compute EER with kaldi
Decision tree
- decision tree balance using 100h Chinese and 20h English --zhiyong
Data selection unsupervised learning
- hold
- acoustic feature based submodular using Pingan dataset --zhiyong
- write code to speed up --zhiyong
- curriculum learning --zhiyong
RNN-DAE(Deep based Auto-Encode-RNN)
- hold
- RNN-DAE has worse performance than DNN-DAE because training dataset is small
- extract real room impulse to generate WSJ reverberation data, and then train RNN-DAE
Ivector&Dvector based ASR
- learning from ivector --Lantian
- CNN ivector learning
- DNN ivector learning
- binary ivector
- metric learning
language vector
- write a paper--zhiyuan
- hold
- language vector is added to multi hidden layers--zhiyuan
- RNN language vector
- hold
multi-GPU
- multi-stream training --Sheng Su
- two GPUs work well, but four GPUs divergent
- kaldi-nnet3 --Xuewei
multi-task
- test according to selt-information neural structure learning --mengyuan
- speech rate learning --xiangyu
=Neutral picture style transfer
- hold
- reproduced the result of the paper "A neutral algorithm of artistic style" --Zhiyuan, Xuewei
- while subject to the GPU's memory, limited to inception net with sgd optimizer (VGG network with the default L-BFGS optimizer consumes very much memory, which is better)
Multi-task learning
- train model using speech rate --xiangyu
- speech recognition plus speaker reconition --xiangyu,lantian,zhiyuan
Text Processing
RNN LM
- character-lm rnn(hold)
- lstm+rnn
- check the lstm-rnnlm code about how to Initialize and update learning rate.(hold)
Neural Based Document Classification
- (hold)
RNN Rank Task
- Test.
- Paper: RNN Rank Net.
- (hold)
- Output rank information.
Graph RNN
- Entity path embeded to entity.
- (hold)
RNN Word Segment
- Set bound to word segment.
- (hold)
Seq to Seq(09-15)
- Review papers.
- Reproduce baseline. (08-03 <--> 08-17)
Order representation
- Nested Dropout
- semi-linear --> neural based auto-encoder.
- modify the objective function(hold)
Balance Representation
- Find error signal
Recommendation
- Reproduce baseline.
- LDA matrix dissovle.
- LDA (Text classification & Recommendation System) --> AAAI
RNN based QA
- Read Source Code.
- Attention based QA.
- Coding.
RNN Poem Process
- Seq based BP.
- (hold)
Text Group Intern Project
Buddhist Process
- (hold)
RNN Poem Process
- Done by Haichao yu & Chaoyuan zuo Mentor : Tianyi Luo.
RNN Document Vector
- (hold)
Image Baseline
- Demo Release.
- Paper Report.
- Read CNN Paper.
Text Intuitive Idea
Trace Learning
- (Hold)
Match RNN
- (Hold)
financial group
model research
- RNN
- online model, update everyday
- modify cost function and learning method
- add more feature
rule combination
- GA method to optimize the model
basic rule
- classical tenth model
multiple-factor
- add more factor
- use sparse model
display
- bug fixed
- buy rule fixed
data
- data api
- download the future data and factor data
