keras-team/keras · error · ValueError

Expected the input image to be rank 3 or 4. Received inputs.

Error message

Expected the input image to be rank 3 or 4. Received inputs.shape={images_shape}

What it means

RandomColorDegeneration.get_random_transformation branches on input rank: 3 = single image, 4 = batch. Any other rank (2, 5, ...) raises this ValueError before computing the degeneration factor.

Source

Thrown at keras/src/layers/preprocessing/image_preprocessing/random_color_degeneration.py:84

            raise ValueError(
                self._VALUE_RANGE_VALIDATION_ERROR
                + f"Received: value_range={value_range}"
            )
        self.value_range = sorted(value_range)

    def get_random_transformation(self, data, training=True, seed=None):
        if isinstance(data, dict):
            images = data["images"]
        else:
            images = data
        images_shape = self.backend.shape(images)
        rank = len(images_shape)
        if rank == 3:
            batch_size = 1
        elif rank == 4:
            batch_size = images_shape[0]
        else:
            raise ValueError(
                "Expected the input image to be rank 3 or 4. Received: "
                f"inputs.shape={images_shape}"
            )

        if seed is None:
            seed = self._get_seed_generator(self.backend._backend)

        factor = self.backend.random.uniform(
            (batch_size, 1, 1, 1),
            minval=self.factor[0],
            maxval=self.factor[1],
            seed=seed,
        )
        factor = factor
        return {"factor": factor}

    def transform_images(self, images, transformation=None, training=True):
        if training:

View on GitHub (pinned to 7a34a03db6)

Solutions

1. Ensure inputs are (H, W, 3) or (batch, H, W, 3)
2. Add the channel axis: images[..., None]
3. Drop extra dims with np.squeeze then verify len(shape)

Example fix

# before
images = gray_array  # (H, W)
out = layer(images)
# after
images = gray_array[..., None]  # (H, W, 1)
out = layer(images)

if len(images.shape) == 2:
    images = images[..., None]

def is_rank3or4(x):
    return len(getattr(x, 'shape', ())) in (3, 4)

When it happens

Trigger: Feeding a (H, W) grayscale array without a channel axis, or a rank-5 tensor with a duplicated batch dim.

Common situations: Grayscale pipelines after convert('L'); over-batched tensors from previous dataset .batch() plus manual expand_dims.

Related errors

• Expected the input image to be rank 3 or 4. Received inputs.
• Invalid images rank: expected rank 3 (single image) or rank
• Expected the input image to be rank 3 or 4. Received inputs.
• {self._VALUE_RANGE_VALIDATION_ERROR}Received: value_range={v
• The `value_range` argument should be a list of two numbers.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/7c5b0071acaecadb. Report an issue: GitHub.