Configurable 3D Scene Synthesis and 2D Image Rendering with Per-pixel Ground Truth Using Stochastic Grammars

This work proposes a systematic learning-based approach to the generation of massive quantities of synthetic 3D scenes and arbitrary numbers of photorealistic 2D images thereof, for the purposes of training, benchmarking, and diagnosing learning-based computer vision and robotics algorithms. In particular, the authors devise a learning-based pipeline of algorithms capable of automatically generating and rendering a potentially infinite variety of indoor scenes by using a stochastic grammar, represented as an attributed Spatial And-Or Graph, in conjunction with state-of-the-art physics-based rendering. The pipeline is capable of synthesizing scene layouts with high diversity, and it is configurable inasmuch as it enables the precise customization and control of important attributes of the generated scenes. It renders photorealistic RGB images of the generated scenes while automatically synthesizing detailed, per-pixel ground truth data, including visible surface depth and normal, object identity, and material information (detailed to object parts), as well as environments (e.g., illuminations and camera viewpoints). The authors demonstrate the value of the synthesized dataset, by improving performance in certain machine-learning-based scene understanding tasks—depth and surface normal prediction, semantic segmentation, reconstruction, etc.---and by providing benchmarks for and diagnostics of trained models by modifying object attributes and scene properties in a controllable manner.