def main(args):
cfg = setup(args)
print("DATASETS: ", cfg.DATASETS.TRAIN)
loader = build_x_train_loader(cfg)
for i, items in tqdm.tqdm(enumerate(loader)):
if i % 128 != 0:
continue
save_dir = Path(cfg.OUTPUT_DIR) / f"{items[0]['image_id']}"
save_dir.mkdir(exist_ok=True, parents=True)
write_jpeg(items[0]['image'], str(save_dir / f'img{i}.jpg'))
for j, mask in enumerate(items[0]['instances'].gt_masks):
Image.fromarray(mask.cpu().numpy()).save(save_dir / f'mask{j}.png')
def test_write_jpeg(img_path, tmpdir):
tmpdir = Path(tmpdir)
img = read_image(img_path)
pil_img = F.to_pil_image(img)
torch_jpeg = str(tmpdir / "torch.jpg")
pil_jpeg = str(tmpdir / "pil.jpg")
write_jpeg(img, torch_jpeg, quality=75)
pil_img.save(pil_jpeg, quality=75)
with open(torch_jpeg, "rb") as f:
torch_bytes = f.read()
with open(pil_jpeg, "rb") as f:
pil_bytes = f.read()
assert_equal(torch_bytes, pil_bytes)
def test_write_jpeg_reference(img_path, tmpdir):
# FIXME: Remove this eventually, see test_encode_jpeg_reference
data = read_file(img_path)
img = decode_jpeg(data)
basedir = os.path.dirname(img_path)
filename, _ = os.path.splitext(os.path.basename(img_path))
torch_jpeg = os.path.join(tmpdir, f"{filename}_torch.jpg")
pil_jpeg = os.path.join(basedir, "jpeg_write", f"{filename}_pil.jpg")
write_jpeg(img, torch_jpeg, quality=75)
with open(torch_jpeg, "rb") as f:
torch_bytes = f.read()
with open(pil_jpeg, "rb") as f:
pil_bytes = f.read()
assert_equal(torch_bytes, pil_bytes)
def test_write_jpeg(img_path):
with get_tmp_dir() as d:
d = Path(d)
img = read_image(img_path)
pil_img = F.to_pil_image(img)
torch_jpeg = str(d / 'torch.jpg')
pil_jpeg = str(d / 'pil.jpg')
write_jpeg(img, torch_jpeg, quality=75)
pil_img.save(pil_jpeg, quality=75)
with open(torch_jpeg, 'rb') as f:
torch_bytes = f.read()
with open(pil_jpeg, 'rb') as f:
pil_bytes = f.read()
assert_equal(torch_bytes, pil_bytes)
def test_write_jpeg(self):
for img_path in get_images(IMAGE_ROOT, ".jpg"):
img = read_jpeg(img_path)
basedir = os.path.dirname(img_path)
filename, _ = os.path.splitext(os.path.basename(img_path))
torch_jpeg = os.path.join(basedir,
'{0}_torch.jpg'.format(filename))
pil_jpeg = os.path.join(basedir, 'jpeg_write',
'{0}_pil.jpg'.format(filename))
write_jpeg(img, torch_jpeg, quality=75)
with open(torch_jpeg, 'rb') as f:
torch_bytes = f.read()
with open(pil_jpeg, 'rb') as f:
pil_bytes = f.read()
os.remove(torch_jpeg)
self.assertEqual(torch_bytes, pil_bytes)
def test(self):
multiplier = 4
path = "/Users/amorvan/Documents/code_dw/neural-styles/images/bw_svg_neural_style/" \
"result_n_paths200_im_size224_n_steps500_layer_nameVGGLayers.Conv4_3_layer_index2.svg"
s = ScaledSvgGen(path, multiplier, multiplier)
f = s.gen_func()
shapes, shape_groups = f()
self.assertEqual(len(shapes), multiplier * multiplier * 200)
self.assertTrue(any(np.max(shape.points.numpy()) < multiplier * 224 for shape in shapes))
render = pydiffvg.RenderFunction.apply
scene_args = pydiffvg.RenderFunction.serialize_scene( \
multiplier * 224, multiplier * 224, shapes, shape_groups)
img = render(multiplier * 224, multiplier * 224, 2, 2, 0,
torch.ones((multiplier * 224, multiplier * 224, 4)), *scene_args)
print(img.shape)
write_jpeg(255 * img.permute(2, 0, 1)[:3, :, :].to(torch.uint8), "test.jpg")
def test_write_jpeg(self):
with get_tmp_dir() as d:
for img_path in get_images(ENCODE_JPEG, ".jpg"):
data = read_file(img_path)
img = decode_jpeg(data)
basedir = os.path.dirname(img_path)
filename, _ = os.path.splitext(os.path.basename(img_path))
torch_jpeg = os.path.join(d, '{0}_torch.jpg'.format(filename))
pil_jpeg = os.path.join(basedir, 'jpeg_write',
'{0}_pil.jpg'.format(filename))
write_jpeg(img, torch_jpeg, quality=75)
with open(torch_jpeg, 'rb') as f:
torch_bytes = f.read()
with open(pil_jpeg, 'rb') as f:
pil_bytes = f.read()
self.assertEqual(torch_bytes, pil_bytes)
def test_write_jpeg_reference(img_path):
# FIXME: Remove this eventually, see test_encode_jpeg_reference
with get_tmp_dir() as d:
data = read_file(img_path)
img = decode_jpeg(data)
basedir = os.path.dirname(img_path)
filename, _ = os.path.splitext(os.path.basename(img_path))
torch_jpeg = os.path.join(d, '{0}_torch.jpg'.format(filename))
pil_jpeg = os.path.join(basedir, 'jpeg_write',
'{0}_pil.jpg'.format(filename))
write_jpeg(img, torch_jpeg, quality=75)
with open(torch_jpeg, 'rb') as f:
torch_bytes = f.read()
with open(pil_jpeg, 'rb') as f:
pil_bytes = f.read()
assert_equal(torch_bytes, pil_bytes)
from torchvision.io.video import read_video, write_video
from torchvision.io.image import write_jpeg
import torch
from args_util import meow_parse
from data_flow import get_predict_video_dataloader
from models import create_model
import os
from visualize_util import save_density_map_normalize, save_density_map
VIDEO_PATH = "/home/tt/Videos/VID_20201204_133931_404.mp4"
OUTPUT_PATH = "/data/my_crowd_image/video_bike_q100"
v, a, info = read_video(VIDEO_PATH, pts_unit='sec')
print(info)
print(v.shape)
length = v.shape[0]
print(length)
count = 0
for i in range(length):
# if (i% 20 == 0):
frame = v[i]
frame = frame.permute(2, 0, 1)
file_out_path = os.path.join(OUTPUT_PATH, "IMG_" + str(i) + ".jpg")
write_jpeg(frame, file_out_path, quality=100)
print(file_out_path)
