def post_process()

in source/neo/eval.py [0:0]

• 28 lines of code
• 7 McCabe index (conditional complexity)

    def post_process(self, class_ids, mx_scores, mx_bounding_boxes):
        """
        rescale the bounding boxes back to its original space

        :param class_ids: ids of detected objects
        :param mx_scores: confidences of bounding boxes
        :param mx_bounding_boxes: bounding boxes of detected objects
        :return: detection response (dictionary)
        """
        if not isinstance(class_ids, np.ndarray):
            class_ids = class_ids.asnumpy()

        if not isinstance(mx_scores, np.ndarray):
            mx_scores = mx_scores.asnumpy()

        if not isinstance(mx_bounding_boxes, np.ndarray):
            mx_bounding_boxes = mx_bounding_boxes.asnumpy()

        # rescale detection results back to original image size
        scale_ratio = self._short_size / self._height if self._height < self._width else self._short_size / self._width
        bbox_coords, bbox_scores = list(), list()
        for index, bbox in enumerate(mx_bounding_boxes[0]):
            prob = float(mx_scores[0][index][0])
            if prob < 0.0:
                continue

            [x_min, y_min, x_max, y_max] = bbox
            x_min = int(x_min / scale_ratio)
            y_min = int(y_min / scale_ratio)
            x_max = int(x_max / scale_ratio)
            y_max = int(y_max / scale_ratio)
            bbox_coords.append([x_min, y_min, x_max, y_max])
            bbox_scores.append([prob])

        body = {
            'width': self._width,
            'height': self._height,
            'channels': self._channels,
            'bbox_scores': bbox_scores,  # shape = (N, 1)
            'bbox_coords': bbox_coords,  # shape = (N, 4)
        }
        return body