Bolei Zhou

Take a look at this recent project from NYU: https://github.com/ai4ce/RealCity3D https://openreview.net/pdf?id=VjFn0L5T7NU. I know the PI and potentially we can have some collaboration

No idea. maybe you could try useGPU = 1. Model is cached using python2.7.

Indeed, it is the pickle library in python which cannot correctly load the CNN model trained in python2.7 into python3.6. Please let me know if there is some good solution...

Due to the upgrading of pytorch0.4, the densenet model structure is also updated, thus the old model cannot be loaded properly. I don't have solution for that yet.

I will add a testing case on the test2015 soon. It would load the trained model then evaluate it on the test2015 data and create the csv which could be...