I am wondering if this CNN is designed just for 2D data like images? Does it reshape features directly into 2D and then start the training?
