#Usage of the web demo

Requirements

The demo server requires Python with some dependencies. To make sure you have the dependencies, please run pip install -r requirements.txt, and also make sure that you have compiled the Python Caffe interface and that it is on your PYTHONPATH

Prepare your caffe model.

Please refer to Caffe for how to train a model based on your own dataset. After the train, you should get the net file .prototxt, the weight file .caffemodel.

Example of how to train the dog breed model based on stanford dog dataset

Plese make sure your caffe framewok was installed, and the path of caffe/.build_release/tools has been added to the system PATH.

1. Download the stanford dog dataset

Download the images, annotations and lists from stanford dogs dataset and extract them to data/standord_dogs

2. Paser the dataset

cd data/standord-dogs && python dog_parse.py

3. Creat the lmdb data for train

cd data/standord-dogs && ./create_image.sh

4. Train the model based on GoogleNet.

Download GoogleNet Model to models/bvlc_googlenet/

Run the following command to start the train.

cd models/bvlc_googlenet/ && train.sh

How to config the web service.

1. You could store your caffe mdoel in the models dir.

2. Modify the web_demo/app.py file.

   a). In line 10, define the path that you store this project. b). define your caffe model path from line 101 t0 108.

http://vision.stanford.edu/aditya86/ImageNetDogs/images.tar http://vision.stanford.edu/aditya86/ImageNetDogs/annotation.tar http://vision.stanford.edu/aditya86/ImageNetDogs/lists.tar

Run

Running python web_demo/app.py will bring up the demo server, accessible at http://0.0.0.0:5000. You can enable debug mode of the web server, or switch to a different port:

% python examples/web_demo/app.py -h
Usage: app.py [options]

Options:
  -h, --help            show this help message and exit
  -d, --debug           enable debug mode
  -p PORT, --port=PORT  which port to serve content on