Spaces:
Runtime error
Runtime error
from tabnanny import filename_only | |
import numpy as np | |
import cv2 | |
import os | |
import PIL | |
import torch | |
from .dataset import Dataset | |
class ImageDataset(Dataset): | |
def __init__(self, | |
img_path, # Path to images. | |
resolution = None, # Ensure specific resolution, None = highest available. | |
**super_kwargs, # Additional arguments for the Dataset base class. | |
): | |
self.sz = resolution | |
self.img_path = img_path | |
self._type = 'dir' | |
self.files = [] | |
self.idx = 0 | |
self._all_fnames = [os.path.relpath(os.path.join(root, fname), start=self.img_path) for root, _dirs, files in os.walk(self.img_path) for fname in files] | |
PIL.Image.init() | |
self._image_fnames = sorted(os.path.join(self.img_path,fname) for fname in self._all_fnames if self._file_ext(fname) in PIL.Image.EXTENSION) | |
if len(self._image_fnames) == 0: | |
raise IOError('No image files found in the specified path') | |
self.files = [] | |
for f in self._image_fnames: | |
if not '_mask' in f: | |
self.files.append(f) | |
self.files = sorted(self.files) | |
def __len__(self): | |
return len(self.files) | |
def _file_ext(fname): | |
return os.path.splitext(fname)[1].lower() | |
def _load_image(self, fn): | |
return PIL.Image.open(fn).convert('RGB') | |
def _get_image(self, idx): | |
# imgfn, seg_map, img_id = self.data_reader.get_image(idx) | |
fname = self.files[idx] | |
ext = self._file_ext(fname) | |
mask = np.array(self._load_image(fname.replace(ext, f'_mask000{ext}')).convert('L')) / 255 | |
rgb = np.array(self._load_image(fname)) # uint8 | |
return rgb, fname.split('/')[-1].replace(ext, ''), mask | |
def __getitem__(self, idx): | |
rgb, fname, mask = self._get_image(idx) # modal, uint8 {0, 1} | |
rgb = rgb.transpose(2,0,1) | |
mask_tensor = torch.from_numpy(mask).to(torch.float32) | |
mask_tensor = mask_tensor.unsqueeze(0) | |
rgb = torch.from_numpy(rgb.astype(np.float32)) | |
rgb = (rgb.to(torch.float32) / 127.5 - 1) | |
rgb_erased = rgb.clone() | |
rgb_erased = rgb_erased * (1 - mask_tensor) # erase rgb | |
rgb_erased = rgb_erased.to(torch.float32) | |
return rgb, rgb_erased, mask_tensor, fname | |
def collate_fn(data): | |
"""Creates mini-batch tensors from the list of images. | |
We should build custom collate_fn rather than using default collate_fn, | |
because merging caption (including padding) is not supported in default. | |
Args: | |
data: list | |
- image: torch tensor of shape (3, 256, 256). | |
Returns: | |
images: torch tensor of shape (batch_size, 3, 256, 256). | |
""" | |
rgbs, rgbs_erased, mask_tensors, fnames = zip(*data) | |
rgbs = list(rgbs) | |
rgbs_erased = list(rgbs_erased) | |
mask_tensors = list(mask_tensors) | |
fnames = list(fnames) | |
return torch.stack(rgbs, dim=0), torch.stack(rgbs_erased, dim=0), torch.stack(mask_tensors, dim=0), fnames | |
def get_loader(img_path, resolution): | |
"""Returns torch.utils.data.DataLoader for custom coco dataset.""" | |
ds = ImageDataset(img_path=img_path, resolution=resolution) | |
data_loader = torch.utils.data.DataLoader(dataset=ds, | |
batch_size=1, | |
shuffle=False, | |
num_workers=1, | |
collate_fn=collate_fn) | |
return data_loader |