SNeS: Learning Probably Symmetric Neural Surfaces from Incomplete Data
نویسندگان
چکیده
We present a method for the accurate 3D reconstruction of partly-symmetric objects. build on strengths recent advances in neural and rendering such as Neural Radiance Fields (NeRF). A major shortcoming approaches is that they fail to reconstruct any part object which not clearly visible training image, often case in-the-wild images videos. When evidence lacking, structural priors symmetry can be used complete missing information. However, exploiting highly non-trivial: while geometry non-reflective materials may symmetric, shadows reflections from ambient scene are symmetric general. To address this, we apply soft constraint material properties, having factored appearance into lighting, albedo colour reflectivity. evaluate our recently introduced CO3D dataset, focusing car category due challenge reconstructing highly-reflective materials. show it unobserved regions with high fidelity render high-quality novel view images.
منابع مشابه
Learning from Incomplete Data
Survey non-response is an important problem in statistics, economics and social sciences. The paper reviews the missing data framework of Little & Rubin [Little and Rubin, 1986]. It presents a survey of techniques to deal with non-response in surveys using a likelihood based approach. The focuses on the case where the probability of a data missing depends on its value. The paper uses the two-st...
متن کاملLearning from Incomplete Data
Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this paper we review the problem of learning from incomplete data from two statistical perspectives|the likelihood-based and the Bayesian. The goal is twofold: to place current neural network approaches to missing data within a statistical framework, and to describe a set of algorith...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Lecture Notes in Computer Science
سال: 2022
ISSN: ['1611-3349', '0302-9743']
DOI: https://doi.org/10.1007/978-3-031-19824-3_22