dev(narugo): add iou check for end-to-end models

    return y

def _yolo_nms(boxes, scores, thresh: float = 0.7) -> List[int]:

def _yolo_nms(boxes, scores, iou_threshold: float = 0.7) -> List[int]:

    """

    Perform Non-Maximum Suppression (NMS) on bounding boxes.

    :type boxes: np.ndarray

    :param scores: Array of confidence scores for each bounding box.

    :type scores: np.ndarray

    :param thresh: IoU threshold for considering boxes as overlapping. Default is 0.7.

    :type thresh: float

    :param iou_threshold: IoU threshold for considering boxes as overlapping. Default is 0.7.

    :type iou_threshold: float

    :return: List of indices of the boxes to keep after NMS.

    :rtype: List[int]

        inter = w * h

        iou = inter / (areas[i] + areas[order[1:]] - inter)

        inds = np.where(iou <= thresh)[0]

        inds = np.where(iou <= iou_threshold)[0]

        order = order[inds + 1]

    return keep

    """

    if output.shape[-1] == 6:  # for end-to-end models like yolov10

        detections = []

        for x0, y0, x1, y1, score, cls in output[output[:, 4] > conf_threshold]:

        output = output[output[:, 4] > conf_threshold]

        selected_idx = _yolo_nms(output[:, :4], output[:, 4])

        for x0, y0, x1, y1, score, cls in output[selected_idx]:

            x0, y0 = _xy_postprocess(x0, y0, old_size, new_size)

            x1, y1 = _xy_postprocess(x1, y1, old_size, new_size)

            detections.append(((x0, y0, x1, y1), labels[int(cls.item())], float(score)))

            return []

        boxes = _yolo_xywh2xyxy(boxes)

        idx = _yolo_nms(boxes, filtered_max_scores, thresh=iou_threshold)

        idx = _yolo_nms(boxes, filtered_max_scores, iou_threshold=iou_threshold)

        boxes, scores = boxes[idx], scores[idx]

        detections = []