examples/onnx/superresolution.py

#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under th

import os
import numpy as np
from PIL import Image
# pip install python-resize-image
from resizeimage import resizeimage

from singa import device
from singa import tensor
from singa import sonnx
import onnx
from utils import download_model, check_exist_or_download

import logging
logging.basicConfig(level=logging.INFO, format='%(asctime)-15s %(message)s')


def preprocess(img):
    img = resizeimage.resize_cover(img, [224, 224], validate=False)
    img_ycbcr = img.convert('YCbCr')
    img_y_0, img_cb, img_cr = img_ycbcr.split()
    img_ndarray = np.asarray(img_y_0)
    img_4 = np.expand_dims(np.expand_dims(img_ndarray, axis=0), axis=0)
    img_5 = img_4.astype(np.float32) / 255.0
    return img_5, img_cb, img_cr


def get_image():
    # download image
    image_url = 'https://s3.amazonaws.com/model-server/inputs/kitten.jpg'
    img = Image.open(check_exist_or_download(image_url))
    return img


class MyModel(sonnx.SONNXModel):

    def __init__(self, onnx_model):
        super(MyModel, self).__init__(onnx_model)

    def forward(self, *x):
        y = super(MyModel, self).forward(*x)
        return y[0]

    def train_one_batch(self, x, y):
        pass


if __name__ == "__main__":

    url = 'https://github.com/onnx/models/raw/master/vision/super_resolution/sub_pixel_cnn_2016/model/super-resolution-10.tar.gz'
    download_dir = '/tmp/'
    model_path = os.path.join(download_dir, 'super_resolution',
                              'super_resolution.onnx')

    logging.info("onnx load model...")
    download_model(url)
    onnx_model = onnx.load(model_path)

    # preprocess
    logging.info("preprocessing...")
    img = get_image()
    img_y, img_cb, img_cr = preprocess(img)
    # sg_ir = sonnx.prepare(onnx_model) # run without graph
    # y = sg_ir.run([img])

    logging.info("model compling...")
    dev = device.create_cuda_gpu()
    x = tensor.PlaceHolder(img_y.shape, device=dev)
    model = MyModel(onnx_model)
    model.compile([x], is_train=False, use_graph=True, sequential=True)

    # inference
    logging.info("model running...")
    x_batch = tensor.Tensor(device=dev, data=img_y)
    img_y = model.forward(x_batch)
    array_img_y = tensor.to_numpy(img_y)
    img_out_y = Image.fromarray(np.uint8((array_img_y[0] * 255.0).clip(0,
                                                                       255)[0]),
                                mode='L')

    # postprocess
    logging.info("postprocessing...")
    final_img = Image.merge("YCbCr", [
        img_out_y,
        img_cb.resize(img_out_y.size, Image.BICUBIC),
        img_cr.resize(img_out_y.size, Image.BICUBIC),
    ]).convert("RGB")
    final_img.show()