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

Unknown iou type {}

Error message

Unknown iou type {}

What it means

CocoEvaluator.prepare() dispatches prediction-preparation based on the COCO iou_type ('bbox', 'segm', 'keypoints'). If iou_type is anything else, it raises ValueError because there is no prepare_for_coco_* method for it. This guards against typos or unsupported task types passed when constructing/using CocoEvaluator.

Source

Thrown at pytorch_object_detection/yolov3_spp/train_utils/coco_eval.py:66

    def accumulate(self):
        for coco_eval in self.coco_eval.values():
            coco_eval.accumulate()

    def summarize(self):
        for iou_type, coco_eval in self.coco_eval.items():
            print("IoU metric: {}".format(iou_type))
            coco_eval.summarize()

    def prepare(self, predictions, iou_type):
        if iou_type == "bbox":
            return self.prepare_for_coco_detection(predictions)
        elif iou_type == "segm":
            return self.prepare_for_coco_segmentation(predictions)
        elif iou_type == "keypoints":
            return self.prepare_for_coco_keypoint(predictions)
        else:
            raise ValueError("Unknown iou type {}".format(iou_type))

    def prepare_for_coco_detection(self, predictions):
        coco_results = []
        for original_id, prediction in predictions.items():
            if len(prediction) == 0:
                continue

            boxes = prediction["boxes"]
            boxes = convert_to_xywh(boxes).tolist()
            scores = prediction["scores"].tolist()
            labels = prediction["labels"].tolist()

            coco_results.extend(
                [
                    {
                        "image_id": original_id,
                        "category_id": labels[k],
                        "bbox": box,

View on GitHub (pinned to 1ec3fe6f37)

Solutions

1. Use one of the supported iou_type values: 'bbox', 'segm', or 'keypoints'.
2. Check the exact strings passed in the iou_types list when constructing CocoEvaluator and fix typos.
3. If you need a new task type, add a prepare_for_coco_<type> method and extend the dispatch chain in prepare().

Example fix

// before
iou_types = ['box']
coco_evaluator = CocoEvaluator(base_dataset, iou_types=iou_types)
// after
iou_types = ['bbox']
coco_evaluator = CocoEvaluator(base_dataset, iou_types=iou_types)

SUPPORTED = {'bbox', 'segm', 'keypoints'}
assert set(iou_types).issubset(SUPPORTED), f"bad iou_type: {iou_types}"

def is_supported_iou_type(t) -> bool:
    return isinstance(t, str) and t in ('bbox', 'segm', 'keypoints')

try:
    evaluator.update(predictions)
except ValueError as e:
    logging.error(f"COCO eval misconfigured: {e}"); raise

When it happens

Trigger: Calling prepare() (indirectly via update()) on a CocoEvaluator that was constructed with an iou_type string other than 'bbox', 'segm', or 'keypoints' — e.g. a typo like 'box' or 'segmentation'.

Common situations: Typo in the iou_type list passed to CocoEvaluator; copying evaluation code from a detection example into a segmentation or keypoint project and editing the string incorrectly; using a custom task type the evaluator doesn't support.

Related errors

• return_layers are not present in model
• illegal stride value.
• The inverted_residual_setting should not be empty.
• replace_stride_with_dilation should be None or a 3-element t
• not support iou_type: {self.iou_type}

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/40f8d275dea538c1. Report an issue: GitHub.