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).

Similar Docs  Excel Report  more

TitleSimilaritySource
None found