WZMIAOMIAO/deep-learning-for-image-processing · error · ValueError

image: {} isn't RGB mode.

Error message

image: {} isn't RGB mode.

What it means

Identical guard to the efficientnetV2 dataset: MyDataSet.__getitem__ in the shufflenet script raises ValueError when a loaded PIL image's mode is not 'RGB'. Grayscale/palette/RGBA images are rejected because the model expects 3-channel inputs and the transform pipeline assumes RGB.

Source

Thrown at pytorch_classification/Test7_shufflenet/my_dataset.py:21

from torch.utils.data import Dataset


class MyDataSet(Dataset):
    """自定义数据集"""

    def __init__(self, images_path: list, images_class: list, transform=None):
        self.images_path = images_path
        self.images_class = images_class
        self.transform = transform

    def __len__(self):
        return len(self.images_path)

    def __getitem__(self, item):
        img = Image.open(self.images_path[item])
        # RGB为彩色图片，L为灰度图片
        if img.mode != 'RGB':
            raise ValueError("image: {} isn't RGB mode.".format(self.images_path[item]))
        label = self.images_class[item]

        if self.transform is not None:
            img = self.transform(img)

        return img, label

    @staticmethod
    def collate_fn(batch):
        # 官方实现的default_collate可以参考
        # https://github.com/pytorch/pytorch/blob/67b7e751e6b5931a9f45274653f4f653a4e6cdf6/torch/utils/data/_utils/collate.py
        images, labels = tuple(zip(*batch))

        images = torch.stack(images, dim=0)
        labels = torch.as_tensor(labels)
        return images, labels

View on GitHub (pinned to 1ec3fe6f37)

Solutions

  1. Convert offending files to RGB on disk with PIL and re-save them.
  2. Auto-normalize in code: replace the raise with img = img.convert('RGB') in __getitem__.
  3. Audit the dataset first: loop over self.images_path and report every file whose Image.open(p).mode != 'RGB'.

Example fix

// before
if img.mode != 'RGB':
    raise ValueError("image: {} isn't RGB mode.".format(self.images_path[item]))
// after
if img.mode != 'RGB':
    img = img.convert('RGB')
Defensive patterns

Strategy: validation

Validate before calling

from PIL import Image
bad = [p for p in dataset.images_path if Image.open(p).mode != 'RGB']
if bad:
    print("Non-RGB images:", bad)
    raise SystemExit(1)

Type guard

def is_rgb(path: str) -> bool:
    return Image.open(path).mode == 'RGB'

Try / catch

try:
    img, label = next(iter(loader))
except ValueError as e:
    if "isn't RGB mode" in str(e):
        dataset.convert_all_to_rgb()
    else:
        raise

Prevention

When it happens

Trigger: Iterating a DataLoader over MyDataSet where any file in the training/validation folder has PIL mode 'L', 'P', or 'RGBA' — e.g. grayscale PNGs, transparent PNGs, or palette GIFs inside the dataset directories.

Common situations: Mixed-format custom datasets (screenshots, downloaded web images with alpha); label/mask images accidentally placed in the training image folders; images converted by some editor to grayscale.

Related errors

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/e974f2978e02965d. Report an issue: GitHub.