The Scene Language: Representing Scenes with Programs, Words, and Embeddings
This addresses the need for robust and automated high-quality 3D/4D scene generation, with potential applications in graphics and AI, though it appears incremental as it builds on pre-trained models and existing renderers.
The paper tackles the problem of representing visual scenes by introducing the Scene Language, which uses programs, words, and embeddings to describe scene structure, semantics, and identity, resulting in higher fidelity and more controllable scene generation compared to existing methods like scene graphs.
We introduce the Scene Language, a visual scene representation that concisely and precisely describes the structure, semantics, and identity of visual scenes. It represents a scene with three key components: a program that specifies the hierarchical and relational structure of entities in the scene, words in natural language that summarize the semantic class of each entity, and embeddings that capture the visual identity of each entity. This representation can be inferred from pre-trained language models via a training-free inference technique, given text or image inputs. The resulting scene can be rendered into images using traditional, neural, or hybrid graphics renderers. Together, this forms a robust, automated system for high-quality 3D and 4D scene generation. Compared with existing representations like scene graphs, our proposed Scene Language generates complex scenes with higher fidelity, while explicitly modeling the scene structures to enable precise control and editing.