Seeing a Rose in Five Thousand Ways

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhang, Yunzhi, Wu, Shangzhe, Snavely, Noah, Wu, Jiajun
Formato: Preprint
Publicado: 2022
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916252915597312
author Zhang, Yunzhi
Wu, Shangzhe
Snavely, Noah
Wu, Jiajun
author_facet Zhang, Yunzhi
Wu, Shangzhe
Snavely, Noah
Wu, Jiajun
contents What is a rose, visually? A rose comprises its intrinsics, including the distribution of geometry, texture, and material specific to its object category. With knowledge of these intrinsic properties, we may render roses of different sizes and shapes, in different poses, and under different lighting conditions. In this work, we build a generative model that learns to capture such object intrinsics from a single image, such as a photo of a bouquet. Such an image includes multiple instances of an object type. These instances all share the same intrinsics, but appear different due to a combination of variance within these intrinsics and differences in extrinsic factors, such as pose and illumination. Experiments show that our model successfully learns object intrinsics (distribution of geometry, texture, and material) for a wide range of objects, each from a single Internet image. Our method achieves superior results on multiple downstream tasks, including intrinsic image decomposition, shape and image generation, view synthesis, and relighting.
format Preprint
id arxiv_https___arxiv_org_abs_2212_04965
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Seeing a Rose in Five Thousand Ways
Zhang, Yunzhi
Wu, Shangzhe
Snavely, Noah
Wu, Jiajun
Computer Vision and Pattern Recognition
What is a rose, visually? A rose comprises its intrinsics, including the distribution of geometry, texture, and material specific to its object category. With knowledge of these intrinsic properties, we may render roses of different sizes and shapes, in different poses, and under different lighting conditions. In this work, we build a generative model that learns to capture such object intrinsics from a single image, such as a photo of a bouquet. Such an image includes multiple instances of an object type. These instances all share the same intrinsics, but appear different due to a combination of variance within these intrinsics and differences in extrinsic factors, such as pose and illumination. Experiments show that our model successfully learns object intrinsics (distribution of geometry, texture, and material) for a wide range of objects, each from a single Internet image. Our method achieves superior results on multiple downstream tasks, including intrinsic image decomposition, shape and image generation, view synthesis, and relighting.
title Seeing a Rose in Five Thousand Ways
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2212.04965