Repository for papers, datasets and methods for automated imagery generation for the built environment.

Our first paper DALLEURBAN is now accepted!

The DALLEURBAN dataset generated by DALLE-2 can be found here. Stable Diffusion dataset can be found here.

To identify the prompt used to generate an image using Stable Diffusion simply do the following. First lookup the filename on the image, divide it by 2 and that's the line number in the text file which contains the prompt for that image.

If you use this dataset or want to cite the paper, please consider using the following citation:

@misc{https://doi.org/10.48550/arxiv.2208.04139, doi = {10.48550/ARXIV.2208.04139},

url = {https://arxiv.org/abs/2208.04139},

author = {Seneviratne, Sachith and Senanayake, Damith and Rasnayaka, Sanka and Vidanaarachchi, Rajith and Thompson, Jason},

keywords = {Computer Vision and Pattern Recognition (cs.CV), Artificial Intelligence (cs.AI), FOS: Computer and information sciences, FOS: Computer and information sciences},

title = {DALLE-URBAN: Capturing the urban design expertise of large text to image transformers},

publisher = {arXiv},

year = {2022},

copyright = {arXiv.org perpetual, non-exclusive license} }