tensorflow/models · error · ValueError

boxes.shape[-1] is {:d}, but must be 4.

Error message

boxes.shape[-1] is {:d}, but must be 4.

What it means

Error "boxes.shape[-1] is {:d}, but must be 4." thrown in tensorflow/models.

Source

Thrown at official/vision/ops/box_ops.py:39

EPSILON = 1e-8
BBOX_XFORM_CLIP = np.log(1000. / 16.)


def yxyx_to_xywh(boxes):
  """Converts boxes from ymin, xmin, ymax, xmax to xmin, ymin, width, height.

  Args:
    boxes: a numpy array whose last dimension is 4 representing the coordinates
      of boxes in ymin, xmin, ymax, xmax order.

  Returns:
    boxes: a numpy array whose shape is the same as `boxes` in new format.

  Raises:
    ValueError: If the last dimension of boxes is not 4.
  """
  if boxes.shape[-1] != 4:
    raise ValueError(
        'boxes.shape[-1] is {:d}, but must be 4.'.format(boxes.shape[-1]))

  boxes_ymin = boxes[..., 0]
  boxes_xmin = boxes[..., 1]
  boxes_width = boxes[..., 3] - boxes[..., 1]
  boxes_height = boxes[..., 2] - boxes[..., 0]
  new_boxes = np.stack(
      [boxes_xmin, boxes_ymin, boxes_width, boxes_height], axis=-1)

  return new_boxes


def yxyx_to_cycxhw(boxes):
  """Converts box corner coordinates to center plus height and width terms.

  Args:
    boxes: a `Tensor` with last dimension of 4, representing the coordinates of
      boxes in ymin, xmin, ymax, xmax order.

View on GitHub (pinned to e006f5f0d5)

Solutions

1. Pass boxes whose last dimension is 4 ([y1, x1, y2, x2]).
2. Slice or re-encode the boxes so the last dimension equals 4.

When it happens

Trigger: Thrown at official/vision/ops/box_ops.py:39 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of tensorflow/models@e006f5f0d5 (2026-08-24). Data as JSON: /api/errors/482efa4a202d0d6b. Report an issue: GitHub.