I am a master student and I want to use mc-cnn as a part of my thesis. I wanna to ask how can I train the mc-cnn with my own dataset and use it in real time camera? Is that the ground truth only need the disparity map(image) or other information(such as dispnoc.bin)?
I am a master student and I want to use mc-cnn as a part of my thesis. I wanna to ask how can I train the mc-cnn with my own dataset and use it in real time camera? Is that the ground truth only need the disparity map(image) or other information(such as dispnoc.bin)?