Hi @wenbowen123 , thank you for your wonderful work! I have successfully deployed the 20-30-48 pre trained weights of Fast Foundation Stereo on Jetson Orin Nano, and the results are good. The disparity map, depth map, and reconstructed point cloud are shown in the following figure. May I ask when you expect to release the training script train.py? I want to use the data collected by my own camera to train the model again in the future. If there is already a release plan or schedule, I would greatly appreciate it.
Thank you.
