When using the BlazeFace Detector on a CUDA device, the error
face_alignment/detection/blazeface/net_blazeface.py", line 288, in _decode_boxes
x_center = raw_boxes[..., 0] / self.x_scale * \
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu!
would be triggered.
This fixes this issue and allows to use the BlazeFace detector on a cuda device.
When using the BlazeFace Detector on a CUDA device, the error
would be triggered.
This fixes this issue and allows to use the BlazeFace detector on a cuda device.