def evaluate(self, net, iteration, noise_func):
avg_psnr = 0.0
for idx in range(len(self.images)):
# We load the original image, between -0.5 and 0.5
orig_img = self.images[idx]
w = orig_img.shape[2]
h = orig_img.shape[1]
noisy_img = noise_func(orig_img)
# We get the predicted image, between -0.5 and 0.5
pred = util.infer_image_no_uint(net, noisy_img)
# Computing the square root error ?
pred255 = util.clip_to_uint8(pred)
orig255 = util.clip_to_uint8(orig_img)
assert (pred255.shape[2] == w and pred255.shape[1] == h)
sqerr = np.square(orig255.astype(np.float32) - pred255.astype(np.float32))
s = np.sum(sqerr)
cur_psnr = 10.0 * np.log10((255 * 255) / (s / (w * h * 3)))
avg_psnr += cur_psnr
# Saving the images
util.save_image(self.submit_config, pred255, "img_{0}_val_{1}_pred.png".format(iteration, idx))
if iteration == 0:
util.save_image(self.submit_config, orig255, "img_{0}_val_{1}_orig.png".format(iteration, idx))
util.save_image(self.submit_config, noisy_img, "img_{0}_val_{1}_noisy.png".format(iteration, idx))
avg_psnr /= len(self.images)
print('Average PSNR: %.2f' % autosummary('PSNR_avg_psnr', avg_psnr))
def evaluate(self, net, iteration, noise_func):
avg_psnr = 0.0
for idx in range(len(self.images)):
orig_img = self.images[idx]
w = orig_img.shape[2]
h = orig_img.shape[1]
noisy_img = noise_func(orig_img)
#pred255 = util.infer_image(net, noisy_img)
pred255 = util.infer_image(net, orig_img) ##edited
orig255 = util.clip_to_uint8(orig_img)
assert (pred255.shape[2] == w and pred255.shape[1] == h)
sqerr = np.square(
orig255.astype(np.float32) - pred255.astype(np.float32))
s = np.sum(sqerr)
cur_psnr = 10.0 * np.log10((255 * 255) / (s / (w * h * 3)))
avg_psnr += cur_psnr
util.save_image(self.submit_config, pred255,
"img_{0}_val_{1}_pred.png".format(iteration, idx))
if iteration == 0:
util.save_image(
self.submit_config, orig_img,
"img_{0}_val_{1}_orig.png".format(iteration, idx))
util.save_image(
self.submit_config, noisy_img,
"img_{0}_val_{1}_noisy.png".format(iteration, idx))
avg_psnr /= len(self.images)
print('Average PSNR: %.2f' % autosummary('PSNR_avg_psnr', avg_psnr))
def evaluate(self, net, iteration, noise_func):
avg_psnr = 0.0
for idx in range(len(self.images)):
orig_img = self.images[idx]
w = orig_img.shape[2]
h = orig_img.shape[1]
noisy_img = noise_func(orig_img)
# infer_image runs the numpy array through the network.
pred255 = util.infer_image(net, noisy_img)
orig255 = util.clip_to_uint8(orig_img)
assert (pred255.shape[2] == w and pred255.shape[1] == h)
sqerr = np.square(
orig255.astype(np.float32) - pred255.astype(np.float32))
s = np.sum(sqerr)
cur_psnr = 10.0 * np.log10((255 * 255) / (s / (w * h * 3)))
avg_psnr += cur_psnr
# Saves the prediction of all images of this iteration in the results directory.
#util.save_image(self.submit_config, pred255, "img_{0}_val_{1}_pred.png".format(iteration, idx))
#if iteration == 0:
#util.save_image(self.submit_config, orig_img, "img_{0}_val_{1}_orig.png".format(iteration, idx))
#util.save_image(self.submit_config, noisy_img, "img_{0}_val_{1}_noisy.png".format(iteration, idx))
avg_psnr /= len(self.images)
print('Average PSNR: %.2f' % autosummary('PSNR_avg_psnr', avg_psnr))
net = util.load_snapshot(args.network_dir + "/network_169000.pickle")
reader = tf.TFRecordReader()
feats = {'shape': tf.FixedLenFeature([3], tf.int64),
'data1': tf.FixedLenFeature([], tf.string),
'data2': tf.FixedLenFeature([], tf.string)}
def _parse_image_function(example_proto):
return tf.parse_single_example(example_proto, feats)
raw_image_dataset = tf.data.TFRecordDataset(args.tf_train)
dataset = raw_image_dataset.map(_parse_image_function)
dat = dataset.make_one_shot_iterator().get_next()
print(dat)
assert(False)
try:
errs = []
while True:
target_img = tf.reshape(tf.decode_raw(dat["data2"], tf.uint8), dat["shape"]).eval()
pred_img = util.infer_image(net, target_img)
target_img = util.clip_to_uint8(np.mean(target_img, axis=0))
pred_img = util.clip_to_uint8(np.mean(pred_img, axis=0))
errs.append(sum(np.sqrt((target_img - pred_img)**2).flatten()))
print(errs)
except tf.errors.OutOfRangeError:
pass