Update modle_arcitecture.md

doc/src/modle_arcitecture.md

@@ -1,25 +1,39 @@
 # Model Arcitecture
 
- The implemented arcitecure of Deepspeech2 online model is based on [Deepspeech2 model](https://arxiv.org/pdf/1512.02595.pdf) with some changes. 
- The model is mainly composed of 2D convolution subsampling layer and single direction rnn layers. 
- To illustrate the model implementation clearly, 5 parts is described in detail.  
-     1. Feature Extraction.
-     2. 2D Convolution subsampling layer.
-     3. RNN layer with only forward direction.
-     4. Softmax Layer.
-     5. CTC Decoder.
+The implemented arcitecure of Deepspeech2 online model is based on [Deepspeech2 model](https://arxiv.org/pdf/1512.02595.pdf) with some changes. 
+The model is mainly composed of 2D convolution subsampling layer and stacked single direction rnn layers. 
+
+To illustrate the model implementation clearly, 3 parts are described in detail.  
+- Feature Extraction.
+- Softmax Layer.
+- Decoder.
+
+
 The arcitecture of the model is shown in Fig.1. 
 
 <p align="center">
 <img src="../images/ds2onlineModel.png" width=800> 
-<br/>Fig.1 The Arcitecture of deepspeech2 online modle
+<br/>Fig.1 The Arcitecture of deepspeech2 online model
 </p>
 
-# Feature Extraction
+# Data Preparation
+## Vocabulary
+For English data, the vocabulary dictionary is composed of 26 English characters with <blank> and <eos>. The <blank> represents the blank label in CTC, the <eos> represents the start and the end of the label. For mandarin, the vocabulary dictionary is composed of chinese characters statisticed from the training set and three additional characters, which are <blank>, <unk> and <eos>. The <unk> represents the unknown characters. 
 
- Three methods of feature extraction is implemented, which are linear, fbank and mfcc.
- For a single utterance $x^i$ sampled from the training set $S$,
+## Mean and Std
+For mean and std, a subset of traininig set is chosed and be used to caculate the mean and std value of the raw audio. 
+ 
+## Feature Extraction
+ For feature extraction, three methods are implemented, which are linear (FFT without using filter bank), fbank and mfcc.
+ Currently, the released deepspeech2 online model use the linear feature extraction method.
+ <!--
+For a single utterance $x^i$ sampled from the training set $S$,
  $ S= {(x^1,y^1),(x^2,y^2),...,(x^m,y^m)}$, where $y^i$ is the label correspodding to the ${x^i}
-
+-->
 # Backbone
-The Backbone is composed of 2D Convolution subsampling layer.
+The Backbone is composed of two 2D convolution subsampling layers and a number of stacked single direction rnn layers. The 2D convolution subsampling layers extract feature represention from the raw audio feature and reduce the length of audio feature at the same time. After passing through the convolution subsampling layers, then the feature represention are input into the stacked rnn layers. For rnn layers, LSTM cell and GRU cell are provided.
+
+# Decoder
+To got the possibilities of characters of each frame, the feature represention of each frame output from the stacked rnn layers are then input into the dense layer to do projection. The output dim of dense layer is same with the vocabulary size. After projection, we use the softmax function to make frame-level feature representation be the possibilities of characters. While making model inference, the character possibilities of each frame are input into the CTC decoder to get the final speech recognition results.
+
+