CaesarNeRF: Calibrated Semantic Representation for Few-shot Generalizable Neural Rendering
This addresses the problem of limited data efficiency in neural rendering for researchers and practitioners, though it appears incremental as it builds on existing NeRF methods with semantic enhancements.
The paper tackles the challenge of achieving generalizability and few-shot learning in Neural Radiance Fields (NeRF) by introducing CaesarNeRF, which uses calibrated semantic representations to improve rendering quality; it demonstrates state-of-the-art performance on datasets like LLFF and MVImgNet, even with a single reference image.
Generalizability and few-shot learning are key challenges in Neural Radiance Fields (NeRF), often due to the lack of a holistic understanding in pixel-level rendering. We introduce CaesarNeRF, an end-to-end approach that leverages scene-level CAlibratEd SemAntic Representation along with pixel-level representations to advance few-shot, generalizable neural rendering, facilitating a holistic understanding without compromising high-quality details. CaesarNeRF explicitly models pose differences of reference views to combine scene-level semantic representations, providing a calibrated holistic understanding. This calibration process aligns various viewpoints with precise location and is further enhanced by sequential refinement to capture varying details. Extensive experiments on public datasets, including LLFF, Shiny, mip-NeRF 360, and MVImgNet, show that CaesarNeRF delivers state-of-the-art performance across varying numbers of reference views, proving effective even with a single reference image.