open-mmlab/mmdetection · error · RuntimeError

results does not contain masks.

Error message

results does not contain masks.

What it means

CopyPaste transform needs source objects; when results lack 'gt_masks' it can only synthesize object masks from boxes if paste_by_box=True. With paste_by_box=False and no gt_masks it raises RuntimeError.

Source

Thrown at mmdet/datasets/transforms/transforms.py:3086

    def get_gt_masks(self, results: dict) -> BitmapMasks:
        """Get gt_masks originally or generated based on bboxes.

        If gt_masks is not contained in results,
        it will be generated based on gt_bboxes.
        Args:
            results (dict): Result dict.
        Returns:
            BitmapMasks: gt_masks, originally or generated based on bboxes.
        """
        if results.get('gt_masks', None) is not None:
            if self.paste_by_box:
                warnings.warn('gt_masks is already contained in results, '
                              'so paste_by_box is disabled.')
            return results['gt_masks']
        else:
            if not self.paste_by_box:
                raise RuntimeError('results does not contain masks.')
            return results['gt_bboxes'].create_masks(results['img'].shape[:2])

    def _select_object(self, results: dict) -> dict:
        """Select some objects from the source results."""
        bboxes = results['gt_bboxes']
        labels = results['gt_bboxes_labels']
        masks = self.get_gt_masks(results)
        ignore_flags = results['gt_ignore_flags']

        selected_inds = self._get_selected_inds(bboxes.shape[0])

        selected_bboxes = bboxes[selected_inds]
        selected_labels = labels[selected_inds]
        selected_masks = masks[selected_inds]
        selected_ignore_flags = ignore_flags[selected_inds]

        results['gt_bboxes'] = selected_bboxes
        results['gt_bboxes_labels'] = selected_labels

View on GitHub (pinned to cfd5d3a985)

Solutions

1. Set paste_by_box=True in the CopyPaste config so masks are created from gt_bboxes
2. Provide gt_masks in the source results (enable with_mask=True in the aux pipeline)
3. Do not use CopyPaste on bbox-only datasets

Example fix

# before
dict(type='CopyPaste', paste_by_box=False)
# after
dict(type='CopyPaste', paste_by_box=True)

assert results.get('gt_masks') is not None or self.paste_by_box, 'need gt_masks or paste_by_box=True'

When it happens

Trigger: Configuring dict(type='CopyPaste', paste_by_box=False) on a dataset/branch that provides only bboxes (no masks), so _get_source_mask has nothing to return.

Common situations: Using CopyPaste on detection-only data; or mis-set paste_by_box when the auxiliary dataset has no mask annotations.

Understand the failure class

Background: "Invalid value" and "allowed values are" config errors: what your library rejected and how to fix it — this error's family across 41 libraries.

Related errors

• Invalid text mode "{self.text_mode}".
• The type of frame_range must be int or list.
• {metric} is not in results
• metric must be a list or a str.
• metric {metric} is not supported.

AI-assisted analysis of open-mmlab/mmdetection@cfd5d3a985 (2026-08-27). Data as JSON: /api/errors/f5579ed3b8416839. Report an issue: GitHub.