This is the source code & repository for the PixBike project for the course of Algorithmic design of Architectural Informatics at TUM. This project aims to create a Pix2Pix model based on the project of GrowBike.
Booklet: PixBike Booklet
Colab: PixBike
You can check the valid training run, the test run for 55th epoch and all the runs can be found in the W&B dashboard.
-
M. Szell, S. Mimar, T. Perlman, G. Ghoshal and R. Sinatra, „Growing urban bicycle networks“, Scientific Reports, vol. 12, no. 1, 2022. Available: https://doi.org/10.1038/s41598-022-10783-y [Accessed 31 August 2022].
-
P. Isola, J. Zhu, T. Zhou and A. Efros, „Image-to-Image Translation with Conditional Adversarial Networks“, CVPR, 2017. Available: https://doi.org/10.48550/arXiv.1611.07004 [Accessed 31 August 2022].