no code implementations • 7 Dec 2023 • Yuto Enyo, Ko Nishino
In this paper, we introduce the first stochastic inverse rendering method, which recovers the attenuated frequency spectrum of an illumination jointly with the reflectance of an object of known geometry from a single image.
no code implementations • 25 Jul 2022 • Kohei Yamashita, Yuto Enyo, Shohei Nobuhara, Ko Nishino
Our key idea is to formulate MVS as an end-to-end learnable network, which we refer to as nLMVS-Net, that seamlessly integrates radiometric cues to leverage surface normals as view-independent surface features for learned cost volume construction and filtering.