torch_npu.npu_ps_roi_pooling(x, rois, spatial_scale, group_size, output_dim) -> Tensor
执行Position Sensitive ROI Pooling。
>>> roi = torch.tensor([[[1], [2], [3], [4], [5]], [[6], [7], [8], [9], [10]]], dtype = torch.float16).npu() >>> x = torch.tensor([[[[ 1]], [[ 2]], [[ 3]], [[ 4]], [[ 5]], [[ 6]], [[ 7]], [[ 8]]], [[[ 9]], [[10]], [[11]], [[12]], [[13]], [[14]], [[15]], [[16]]]], dtype = torch.float16).npu() >>> out = torch_npu.npu_ps_roi_pooling(x, roi, 0.5, 2, 2) >>> outtensor([[[[0., 0.], [0., 0.]], [[0., 0.], [0., 0.]]], [[[0., 0.], [0., 0.]], [[0., 0.], [0., 0.]]]], device='npu:0', dtype=torch.float16)