adjust picture size

README.md

@@ -1,20 +1,27 @@
 # TextSnake: A Flexible Representation for Detecting Text of Arbitrary Shapes
 
-A PyTorch implement of **TextSnake: A Flexible Representation for Detecting Text of Arbitrary Shapes** (Face++ ECCV 2018)
+A PyTorch implement of **TextSnake: A Flexible Representation for Detecting Text of Arbitrary Shapes** (ECCV 2018) by `Face++`
 
 - Paper link: [arXiv:1807.01544](https://arxiv.org/abs/1807.01544)
 
 - Github: [princewang1994/TextSnake.pytorch](https://github.com/princewang1994/TextSnake.pytorch)
 
-![](http://princepicbed.oss-cn-beijing.aliyuncs.com/blog_20181228201251.png)
+<div style="color:#0000FF" align="center">
+<img src="http://princepicbed.oss-cn-beijing.aliyuncs.com/blog_20181228201251.png" width="630"/> 
+</div>
+
 
 ## Paper
 
+
 ![](http://princepicbed.oss-cn-beijing.aliyuncs.com/blog_20181228172334.png)
 
 Comparison of different representations for text instances. (a) Axis-aligned rectangle. (b) Rotated rectangle. (c) Quadrangle. (d) TextSnake. Obviously, the proposed TextSnake representation is able to effectively and precisely describe the geometric properties, such as location, scale, and bending of curved text with perspective distortion, while the other representations (axis-aligned rectangle, rotated rectangle or quadrangle) struggle with giving accurate predictions in such cases.
 
-![](http://princepicbed.oss-cn-beijing.aliyuncs.com/blog_20181228172346.png)
+<div style="color:#0000FF" align="center">
+<img src="http://princepicbed.oss-cn-beijing.aliyuncs.com/blog_20181228172346.png" width="530"/> 
+</div>
+
 
 Text snake element:
 
@@ -40,32 +47,32 @@ This repo includes the training code and inference demo of TextSnake, training a
 
 To run this repo successfully, it is highly recommanded with:
 
-- Linux (Ubuntu 16.04)
-- Python3.6
-- Anaconda3
-- NVIDIA GPU(with 8G or larger GPU memroy)
+- **Linux (Ubuntu 16.04)**
+- **Python3.6**
+- **Anaconda3**
+- **NVIDIA GPU(with 8G or larger GPU memroy)**
 
 (I haven't test it on other Python version.)
 
 1. clone this repository
 
-```
+```shell
 git clone https://github.com/princewang1994/TextSnake.pytorch.git
 ```
 
 2. python package can be installed with `pip`
-```
-cd $TEXTSNAKE
-pip install -r requirements.txt
+```shell
+$ cd $TEXTSNAKE
+$ pip install -r requirements.txt
 ```
 
 ## Training
 
 Training model with given experiment name `$EXPNAME`
 
 ```shell
-EXPNAME=example
-CUDA_VISIBLE_DEVICES=$GPUID python train.py $EXPNAME --viz
+$ EXPNAME=example
+$ CUDA_VISIBLE_DEVICES=$GPUID python train.py $EXPNAME --viz
 ```
 
 **options:**
@@ -80,8 +87,8 @@ other options can be show by run `python train.py -h`
 Runing following command can generate demo on TotalText dataset (300 pictures), the result are save to `./vis` by default
 
 ```shell
-EXPNAME=example
-CUDA_VISIBLE_DEVICES=$GPUID python demo.py --checkepoch 190
+$ EXPNAME=example
+$ CUDA_VISIBLE_DEVICES=$GPUID python demo.py --checkepoch 190
 ```
 
 **options:**
@@ -92,7 +99,9 @@ other options can be show by run `python train.py -h`
 
 ## Performance
 
-left: prediction, middle: text region(TR), right: text center line(TCL)
+- left: prediction
+- middle: text region(TR)
+- right: text center line(TCL)
 
 ![](demo/24_img650.jpg)
 
@@ -104,7 +113,7 @@ left: prediction, middle: text region(TR), right: text center line(TCL)
 
 ## What is comming
 
-- [ ] more dataset suport: ICDAR15/[SynthText](https://github.com/ankush-me/SynthText)
+- [ ] more dataset suport: [ICDAR15/[SynthText](https://github.com/ankush-me/SynthText)]
 - [ ] Metric computing
 - [ ] Cython/C++ accelerate core functions
 
@@ -114,4 +123,4 @@ This project is licensed under the MIT License - see the [LICENSE.md](LICENSE.md
 
 ## Acknowledgement
 
-This project is writen by [Prince Wang](https://github.com/princewang1994), some code  refer to [songdejia/EAST](https://github.com/songdejia/EAST)
+This project is writen by [Prince Wang](https://github.com/princewang1994), part of codes refer to [songdejia/EAST](https://github.com/songdejia/EAST)