keras-team/keras · error · ValueError

`encoding_format` should be one of 'center_xywh' or 'center_

Error message

`encoding_format` should be one of 'center_xywh' or 'center_yxhw', got {encoding_format}

What it means

encode_box_to_deltas supports two delta encoding formats: 'center_xywh' (center x, y, width, height) and 'center_yxhw' (center y, x, height, width). Any other encoding_format string is rejected with this ValueError before any computation happens.

Source

Thrown at keras/src/layers/preprocessing/image_preprocessing/bounding_boxes/converters.py:308

            `image_shape` is used for normalization.
    Returns:
        Encoded box deltas. The return type matches the `encode_format`.

    Raises:
        ValueError: If `variance` is not None and its length is not 4.
        ValueError: If `encoding_format` is not `"center_xywh"` or
            `"center_yxhw"`.

    """
    if variance is not None:
        variance = ops.convert_to_tensor(variance, "float32")
        var_len = variance.shape[-1]

        if var_len != 4:
            raise ValueError(f"`variance` must be length 4, got {variance}")

    if encoding_format not in ["center_xywh", "center_yxhw"]:
        raise ValueError(
            "`encoding_format` should be one of 'center_xywh' or "
            f"'center_yxhw', got {encoding_format}"
        )

    if image_shape is None:
        height, width = None, None
    else:
        height, width, _ = image_shape

    encoded_anchors = convert_format(
        anchors,
        source=anchor_format,
        target=encoding_format,
        height=height,
        width=width,
    )
    boxes = convert_format(
        boxes,

View on GitHub (pinned to 7a34a03db6)

Solutions

1. Use encoding_format='center_xywh' or 'center_yxhw'.
2. If boxes are in corner format, convert them first, or use a different converter suited to xyxy.

Example fix

# before
deltas = encode_box_to_deltas(boxes, anchors, encoding_format="xywh")
# after
deltas = encode_box_to_deltas(boxes, anchors, encoding_format="center_xywh")

assert encoding_format in ("center_xywh", "center_yxhw"), f"bad encoding_format {encoding_format!r}"

def is_valid_encoding_format(f):
    return f in {"center_xywh", "center_yxhw"}

When it happens

Trigger: Passing encoding_format='xyxy', 'corner', 'centroid', or a typo like 'center_xyw'.

Common situations: Assuming the encoder accepts the same format names as other converters in the module (e.g. 'centers' or 'xywh'); mixing up this function's format names with those of other converters.

Related errors

• `variance` must be length 4, got {variance}
• `encoded_format` should be 'center_xywh' or 'center_yxhw', b
• Cannot concatenate features because feature '{name}' has not
• `height` and `width` must be set if `format='xyxy'`.
• compute_iou() expects boxes1 to be batched, or to be unbatch

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/44f136cd12617b76. Report an issue: GitHub.