I-Scene: 3D Instance Models are Implicit Generalizable Spatial Learners
This work addresses the problem of limited generalization in 3D scene generation for interactive applications, offering a novel approach that is incremental in leveraging existing models.
The paper tackles the challenge of generalization in interactive 3D scene generation by reprogramming a pre-trained 3D instance generator to learn spatial relations from geometric cues, enabling generalization to unseen layouts and novel object compositions without dataset-bounded supervision.
Generalization remains the central challenge for interactive 3D scene generation. Existing learning-based approaches ground spatial understanding in limited scene dataset, restricting generalization to new layouts. We instead reprogram a pre-trained 3D instance generator to act as a scene level learner, replacing dataset-bounded supervision with model-centric spatial supervision. This reprogramming unlocks the generator transferable spatial knowledge, enabling generalization to unseen layouts and novel object compositions. Remarkably, spatial reasoning still emerges even when the training scenes are randomly composed objects. This demonstrates that the generator's transferable scene prior provides a rich learning signal for inferring proximity, support, and symmetry from purely geometric cues. Replacing widely used canonical space, we instantiate this insight with a view-centric formulation of the scene space, yielding a fully feed-forward, generalizable scene generator that learns spatial relations directly from the instance model. Quantitative and qualitative results show that a 3D instance generator is an implicit spatial learner and reasoner, pointing toward foundation models for interactive 3D scene understanding and generation. Project page: https://luling06.github.io/I-Scene-project/