Learning Object-Centric Representations of Multi-Object Scenes from Multiple Views
–Neural Information Processing Systems
As a result, these "single-view" methods form their representations of a 3D scene based only on a single 2D observation (view).
Neural Information Processing Systems
Oct-2-2025, 17:51:45 GMT
- Country:
- North America > Canada (0.04)
- Technology:
- Information Technology > Artificial Intelligence
- Machine Learning (1.00)
- Representation & Reasoning (0.94)
- Vision (1.00)
- Information Technology > Artificial Intelligence