dgs.models.dataset.keypoint_rcnn.KeypointRCNNVideoBackbone.arbitrary_to_ds¶

KeypointRCNNVideoBackbone.arbitrary_to_ds(a: torchvision.tv_tensors.Image | torch.Tensor, idx: int) → list[State][source]¶: Given a frame of a video, return the resulting state after running the RCNN model.