DeepView: View Synthesis with Learned Gradient Descent
This work addresses the problem of generating realistic novel views for computer vision applications, representing an incremental improvement over existing methods.
The paper tackles view synthesis from sparse camera viewpoints using multiplane images and learned gradient descent, achieving state-of-the-art results on the Kalantari and Spaces datasets with improved handling of occlusions and complex scene features.
We present a novel approach to view synthesis using multiplane images (MPIs). Building on recent advances in learned gradient descent, our algorithm generates an MPI from a set of sparse camera viewpoints. The resulting method incorporates occlusion reasoning, improving performance on challenging scene features such as object boundaries, lighting reflections, thin structures, and scenes with high depth complexity. We show that our method achieves high-quality, state-of-the-art results on two datasets: the Kalantari light field dataset, and a new camera array dataset, Spaces, which we make publicly available.