tensorflow/models · error · ValueError

Number of sides is incorrect.

Error message

Number of sides is incorrect.

What it means

Error "Number of sides is incorrect." thrown in tensorflow/models.

Source

Thrown at official/vision/modeling/layers/detection_generator.py:649

      `valid_detections` boxes are valid detections.

  Raises:
    ValueError if inputs shapes are not valid.
  """
  one = tf.constant(1, dtype=scores.dtype)
  with tf.name_scope('generate_detections'):
    batch_size, num_box_classes, box_locations, sides = (
        boxes.get_shape().as_list()
    )
    if batch_size is None:
      batch_size = tf.shape(boxes)[0]
    _, num_classes, locations = scores.get_shape().as_list()
    if num_box_classes != 1 and num_box_classes != num_classes:
      raise ValueError('Boxes should have either 1 class or same as scores.')
    if locations != box_locations:
      raise ValueError('Number of locations is different.')
    if sides != 4:
      raise ValueError('Number of sides is incorrect.')
    # Selects pre_nms_score_threshold scores before NMS.
    boxes, scores = box_ops.filter_boxes_by_scores(
        boxes, scores, min_score_threshold=pre_nms_score_threshold
    )

    # EdgeTPU-friendly class-wise NMS, -1 for invalid.
    indices = edgetpu.non_max_suppression_padded(
        boxes,
        scores,
        max_num_detections,
        iou_threshold=nms_iou_threshold,
        refinements=refinements,
    )
    # Gather NMS-ed boxes and scores.
    safe_indices = tf.nn.relu(indices)  # 0 for invalid
    invalid_detections = safe_indices - indices  # 1 for invalid, 0 for valid
    valid_detections = one - invalid_detections  # 0 for invalid, 1 for valid
    safe_indices = tf.cast(safe_indices, tf.int32)

View on GitHub (pinned to e006f5f0d5)

Solutions

1. Make the last dimension of boxes equal to 4 times the number of classes (or 4 for class-agnostic boxes).
2. Check the box encoding: each box needs exactly 4 coordinates [y1, x1, y2, x2].

When it happens

Trigger: Thrown at official/vision/modeling/layers/detection_generator.py:649 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/cf32216f1fc685da. Report an issue: GitHub.