Compact and adaptive multiplane images for view synthesis
This work addresses a practical bottleneck for applications using MPIs by reducing memory usage, though it appears incremental as it builds on existing MPI methods.
The paper tackles the high memory requirement of Multiplane Images (MPIs) for view synthesis by proposing a learning method that optimizes memory usage to create compact and adaptive MPIs, avoiding redundant information and considering scene geometry for depth sampling.
Recently, learning methods have been designed to create Multiplane Images (MPIs) for view synthesis. While MPIs are extremely powerful and facilitate high quality renderings, a great amount of memory is required, making them impractical for many applications. In this paper, we propose a learning method that optimizes the available memory to render compact and adaptive MPIs. Our MPIs avoid redundant information and take into account the scene geometry to determine the depth sampling.