This is the supplemental material for CVPR 2019 submission 2950, "Fast and Flexible Indoor Scene Synthesis via Deep Convolutional Generative Models."

This archive contains the following:

- supplemental.pdf: Text document containing additional results, details of datasets and model architectures, etc.

- mturk_raw_results: The raw responses, in .csv format, of the participants in our Amazon Mechanical Turk perceptual study.

- mitsuba_renderings: high-quality, physically-based renderings of all the scenes generated by our method which were used in the perceptual study.

For more materials, including higher resolution renderings, as well as all the images used in the perceptual study, visit the github repository of the project: https://github.com/brownvc/fast-synth
