Get mAP result

Get the valid radios of feature maps of all level.

Get weights path from WEIGHTS_HOME, if not exists, download it from url.

Calculate iou of box1 and box2 Args: box1 (Tensor): box with the shape [M1, 4] box2 (Tensor): box with the shape [M2, 4] Ret

Whether path is URL. Args: path (string): URL string or not.

Applies mask to bbox region in image then adds content_tensor to it.

Apply NMS to avoid detecting too many overlapping bounding boxes. Args: dets: shape [N, 5], [score, x1, y1, x2, y2] m

Paste the mask prediction to the original image.

print function help

Return correct prediction matrix Arguments: detections (array[N, 6]), x1, y1, x2, y2, conf, class labels (array[M, 5]), class

It takes the output of the mask head, and applies the mask to the bounding boxes. This produces masks of higher quality but is slower. A

Resize pos_embed weight. Args: pos_embed (Tensor): the pos_embed weight old_hw (list[int]): the height and wi

Applies Shear Transformation to the image and shifts the bboxes. Args: image: 3D uint8 Tensor. bboxes: 2D Tensor that is a list o

Equivalent of PIL Translate in X/Y dimension that shifts image and bbox. Args: image: 3D uint8 Tensor. bboxes: 2D Tensor that is

`Varifocal Loss <https://arxiv.org/abs/2008.13367>`_ Args: pred (Tensor): The prediction with shape (N, C), C is the number o

Args: im (str/np.ndarray): path of image/np.ndarray read by cv2 results (dict): include 'boxes': np.ndarray: shape:[N,6], N: numb

Visualize bbox and mask results

Convert bounding box coordinates from (x1, y1, x2, y2) format to (x, y, width, height) format where (x1, y1) is the top-left corner and (x2,