Intrinsicnerf: Learning intrinsic neural radiance fields for editable novel view synthesis

W Ye, S Chen, C Bao, H Bao… - Proceedings of the …, 2023 - openaccess.thecvf.com
Existing inverse rendering combined with neural rendering methods can only perform
editable novel view synthesis on object-specific scenes, while we present intrinsic neural
radiance fields, dubbed IntrinsicNeRF, which introduce intrinsic decomposition into the
NeRF-based neural rendering method and can extend its application to room-scale scenes.
Since intrinsic decomposition is a fundamentally under-constrained inverse problem, we
propose a novel distance-aware point sampling and adaptive reflectance iterative clustering …

[PDF][PDF] IntrinsicNeRF: Learning Intrinsic Neural Radiance Fields for Editable Novel View Synthesis Supplementary Document

W Ye, S Chen, C Bao, H Bao, M Pollefeys, Z Cui… - openaccess.thecvf.com
In this supplementary document, we provide the framework of the semantic branch in
IntrinsicNeRF (Sec. A), and more experimental results (Sec. B) such as qualitative and
quantitative results on the Blender Object dataset (Sec. B. 1) and the Replica Scene dataset
(Sec. B. 2), and ablation studies (Sec. B. 3). We also present the applicability of our method
on both synthetic and real-world data (Sec. B. 4).
Showing the best results for this search. See all results