Preprocessing video for MTTS-CAN

[JVD9kh96]

Hello, first of all, thanks for this repository, I found it really helpful. Then, I had a question about the video preprocessing for MTTS-CAN. As I noticed, that in the official implementation of this network, the frames are rotated by 90 degrees. But I noticed that this line is removed from your code, or at least, I didn't find it in your pyVHR_demo_deep notebook. In other words, the frames are extracted from a video by taking video path as input, then converting the frames from BGR 2 RGB format, same as the original implementation. The rest of the preprocessing step is pretty similar to the original implementation except that rotation that I mentioned earlier. Is there a particular reason reason for that?
Thanks in advance.