In class LSSViewTransformer -> def get_lidar_coor(...), the camera intrinsic matrixes cam2imgs is of the src size (900, 1600), not of the input size (256, 704).
Reproduction
Add the following code in class LSSViewTransformer -> def get_lidar_coor(...).
print(self.frustum.shape)
print(self.frustum[0, -1, -1, ...].to(int))
print(cam2imgs.to(int))
quit()
Run the test.
python3 tools/test.py configs/bevdet/bevdet-r50.py ../bevdet-r50.pth --format-only --eval-options jsonfile_prefix=out
It shows that the (u, v) of input size (256, 704) will be operated with the camera intrinsic matrix of src size (900, 1600).
It looks like an issue, or is there any special consideration here?
In class LSSViewTransformer -> def get_lidar_coor(...), the camera intrinsic matrixes cam2imgs is of the src size (900, 1600), not of the input size (256, 704).
Reproduction
Add the following code in class LSSViewTransformer -> def get_lidar_coor(...). print(self.frustum.shape) print(self.frustum[0, -1, -1, ...].to(int)) print(cam2imgs.to(int)) quit()
Run the test. python3 tools/test.py configs/bevdet/bevdet-r50.py ../bevdet-r50.pth --format-only --eval-options jsonfile_prefix=out
Get the log. torch.Size([59, 16, 44, 3]) tensor([703, 255, 1]) tensor([[[[1257, 0, 827], [ 0, 1257, 450], [ 0, 0, 1]],
It shows that the (u, v) of input size (256, 704) will be operated with the camera intrinsic matrix of src size (900, 1600). It looks like an issue, or is there any special consideration here?