Robust Photometric Stereo via Low-Rank Matrix Completion and Recovery

By Lun Wu, Arvind Ganesh, Boxin Shi, Yasuyuki Matsushita, Yongtian Wang, Yi Ma. Asian Conference on Computer Vision, 2010

344 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Adaptive 3D Face Reconstruction from Unconstrained Photo Collections (2017, background): “Later it was discovered that even without detailed knowledge of the light source, photometric stereo can take advantage of dominant contribution of the low-rank spherical harmonics in lighting due to the mainly diffuse reflectance of face [40], [41], [42], [43], [44], [45].”
Practical low-rank matrix approximation under robust L1-norm (2012, methodology): “Links to trace-norm relaxation: Our regularizer appears similar to the trace-norm relaxation term in [9, 18], but the essence is quite different.”
Unifying Nuclear Norm and Bilinear Factorization Approaches for Low-Rank Matrix Decomposition (2013, neutral): “828 [38] 33.”
Toward Guaranteed Illumination Models for Non-convex Objects (2013, background): “The low-rank and sparse decomposition leverages our qualitative understanding of the physical properties of images (low-dimensionality, sparsity of cast shadows) [1, 15, 21, 4, 22], while the constraint ensures that the output of this algorithm is always a good approximation to the target cone.”