Equivariant Neural Rendering
This addresses scene representation and rendering for computer vision applications, offering a novel approach but with incremental gains in speed.
The paper tackles the problem of learning neural scene representations from images without 3D supervision by enforcing equivariance to 3D transformations, achieving real-time inference with results comparable to slower models.
We propose a framework for learning neural scene representations directly from images, without 3D supervision. Our key insight is that 3D structure can be imposed by ensuring that the learned representation transforms like a real 3D scene. Specifically, we introduce a loss which enforces equivariance of the scene representation with respect to 3D transformations. Our formulation allows us to infer and render scenes in real time while achieving comparable results to models requiring minutes for inference. In addition, we introduce two challenging new datasets for scene representation and neural rendering, including scenes with complex lighting and backgrounds. Through experiments, we show that our model achieves compelling results on these datasets as well as on standard ShapeNet benchmarks.