trimscash/mist-server

mist-server

#.env
MIST_DIRECTORY="{mist repository directory}"
ALLOW_ORIGIN="{allow origin url}"

mist/ is sourced from https://github.com/mist-project/mist

change tail of mist_v3.py

# tail of mist_v3.py
        output = Image.fromarray(output_image.astype(np.uint8))
        output_name = args.output_name
        # save_parameter = '_' + str(epsilon) + '_' + str(steps) + '_' + str(input_size) + '_' + str(block_num) + '_' + str(mode) + '_' + str(args.rate) + '_' + str(int(mask)) + '_' + str(int(resize))
        # output_name += save_parameter + '.png'
        # output_name += '.png'
        print("Output image saved in path {}".format(output_name))
        output.save(output_name)

and make directory

mist/temp/input
mist/temp/output

npm install
conda activate mist
npm start

usage

post /

post base64 decoded image with mime

payload

• image: base64 decoded image with mime

#payload example
{ "image" : "data:image/png;base64,~~~~~~~~~"}

response

• image: base64 decoded image with mime

#response example
{ "status":"success", "image" : "data:image/png;base64,~~~~~~~~~"}

curl

curl -X POST http://localhost:4000 -H "Accept: application/json" -H "Content-type: application/json" -d '{ "image" : "data:image/png;base64,~~~~~~~~~"}':

get /

get queue status

response

• active_job indicates the processing job count
• wait_job indicates the waiting job count

#response example
{"status":"success","message":"now queue state","active_job":1,"wait_job":5}