SceneExpander: Real-Time Scene Synthesis for Interactive Floor Plan Editing

Shao-Kui Zhang, Junkai Huang, Liang Yue, Jia-Tong Zhang, Jia-Hong Liu, Yu-Kun Lai, Song-Hai Zhang

Abstract: Scene synthesis has gained significant attention recently, and interactive scene synthesis focuses on yielding scenes according to user preferences. Existing literature either generates floor plans or scenes according to the floor plans. This paper generates scenes over floor plans in real time. Given an initial scene, the only interaction a user needs is changing the room shapes. Our framework splits/merges rooms and adds/rearranges/removes objects for each transient moment during interactions. A systematic pipeline achieves our framework by compressing objects' arrangements over modified room shapes in a transient moment, thus enabling real-time performances. We also propose elastic boxes that indicate how objects should be arranged according to their continuously changed contexts, such as room shapes and other objects. Through a few interactions, a floor plan filled with object layouts is generated concerning user preferences on floor plans and object layouts according to floor plans. Experiments show that our framework is efficient at user interactions and plausible for synthesizing 3D scenes.

Primary Subject Area: [Experience] Multimedia Applications

Secondary Subject Area: [Experience] Art and Culture

Relevance To Conference:

This paper proposes a framework for interactively synthesizing 3D scenes.

First, it aligns with Multimedia Applications in ACM MM. Our framework improves user experiences in arranging 3D objects and editing floor plans. Instead of directly interacting with 3D objects like existing literature, our framework indirectly arranges objects over editing floor plans, i.e., the floor plans are the interacting units controlled by users. To our knowledge, this is the first interactive scene synthesis framework that arranges objects through continuously changing floor plans. Experiments show that our framework is more efficient than existing literature on synthesizing scenes.

Second, as a scene synthesis method, we propose "elastic boxes" that encode how objects are arranged concerning walls, which aligns with Art and Culture in ACM MM. Our elastic boxes are generative because they directly yield arrangements for a group of objects. Our elastic boxes are also dynamic because the walls can be continuously changed, resulting in continuously rearranging the objects. Finally, our framework can be recognized as a computational tool for interior designers or normal users to create 3D scenes efficiently.

Submission Number: 24