site stats
Photorealistic Scene Synthesis


We present a human-centric method to sample and synthesize 3D room layouts and 2D images thereof, for the purpose of obtaining large-scale 2D/3D image data with the perfect per-pixel ground truth. An attributed spatial And-Or graph (S-AOG) is proposed to represent indoor scenes. The S-AOG is a probabilistic grammar model, in which the terminal nodes are object entities including room, furniture, and supported objects. Human contexts as contextual relations are encoded by Markov Random Fields (MRF) on the terminal nodes. We learn the distributions from an indoor scene dataset and sample new layouts using Monte Carlo Markov Chain. Experiments demonstrate that the proposed method can robustly sample a large variety of realistic room layouts based on three criteria: (i) visual realism comparing to a state-of-the-art room arrangement method, (ii) accuracy of the affordance maps with respect to ground-truth, and (ii) the functionality and naturalness of synthesized rooms evaluated by human subjects.


Please cite our paper if you use our code or data.

    title={Human-centric Indoor Scene Synthesis Using Stochastic Grammar},
    author={Qi, Siyuan and Zhu, Yixin and Huang, Siyuan and Jiang, Chenfanfu and Zhu, Song-Chun},
    booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},

The authors thank Professor Ying Nian Wu from UCLA Statistics Department and Professor Demetri Terzopoulos from UCLA Computer Science Department for insightful discussions. The work reported herein is supported by DARPA XAI N66001-17-2-4029 and ONR MURI N00014-16-1-2007.