KeyPoints

class torchvision.tv_tensors.KeyPoints(data: Any, *, canvas_size: tuple[int, int], dtype: Optional[dtype] = None, device: Optional[Union[device, str, int]] = None, requires_grad: Optional[bool] = None)

torch.Tensor subclass for tensors with shape [..., 2] that represent points in an image.

Note

Support for keypoints was released in TorchVision 0.23 and is currently a BETA feature. We don’t expect the API to change, but there may be some rare edge-cases. If you find any issues, please report them on our bug tracker: https://github.com/pytorch/vision/issues?q=is:open+is:issue Each point is represented by its X and Y coordinates along the width and height dimensions, respectively.

Each point is represented by its X and Y coordinates along the width and height dimensions, respectively.

KeyPoints may represent any object that can be represented by sequences of 2D points:

• Polygonal chains, including polylines, Bézier curves, etc., which can be of shape [N_chains, N_points, 2].

• Polygons, which can be of shape [N_polygons, N_points, 2].

• Skeletons, which can be of shape [N_skeletons, N_bones, 2, 2] for pose-estimation models.

Note

Like for torchvision.tv_tensors.BoundingBoxes, there should only be a single instance of the torchvision.tv_tensors.KeyPoints class per sample e.g. {"img": img, "poins_of_interest": KeyPoints(...)}, although one torchvision.tv_tensors.KeyPoints object can contain multiple key points

Parameters:

• data – Any data that can be turned into a tensor with torch.as_tensor().

• canvas_size (two-tuple of python:ints) – Height and width of the corresponding image or video.

• dtype (torch.dpython:type, optional) – Desired data type of the bounding box. If omitted, will be inferred from data.

• device (torch.device, optional) – Desired device of the bounding box. If omitted and data is a torch.Tensor, the device is taken from it. Otherwise, the bounding box is constructed on the CPU.

• requires_grad (bool, optional) – Whether autograd should record operations on the bounding box. If omitted and data is a torch.Tensor, the value is taken from it. Otherwise, defaults to False.

Examples using KeyPoints:

Getting started with transforms v2

Getting started with transforms v2

TVTensors FAQ

TVTensors FAQ