2014 IEEE International Symposium on Mixed and Augmented Reality (ISMAR) 2014
DOI: 10.1109/ismar.2014.6948416
|View full text |Cite
|
Sign up to set email alerts
|

Real-time illumination estimation from faces for coherent rendering

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
38
0
1

Year Published

2015
2015
2021
2021

Publication Types

Select...
4
3
1

Relationship

0
8

Authors

Journals

citations
Cited by 25 publications
(39 citation statements)
references
References 28 publications
0
38
0
1
Order By: Relevance
“…The approach of Knorr and Kurz [15] estimates incident light from human faces using machine learning from a variety of faces under different illumination conditions. The method applies a face tracker and matches distinctive observation points of the detected face to the training database to estimate the real-world light.…”
Section: Related Workmentioning
confidence: 99%
See 2 more Smart Citations
“…The approach of Knorr and Kurz [15] estimates incident light from human faces using machine learning from a variety of faces under different illumination conditions. The method applies a face tracker and matches distinctive observation points of the detected face to the training database to estimate the real-world light.…”
Section: Related Workmentioning
confidence: 99%
“…We empirically found a space of 2 15 illumination variations sufficient, which is created by setting the ambient light C 0 = 1 and quantizing the higher-order SH coefficients C 1 to C 15 with two different values in the range of -1 to 1. Note, selecting only two different values for each coefficient keeps the size of the database small enough for training the CNNs in an acceptable amount of time.…”
Section: Synthesis Of Training Imagesmentioning
confidence: 99%
See 1 more Smart Citation
“…However these approaches cannot handle inserting a specific model, but rather proposes a set of images of objects in a specific category that roughly match illumination conditions in the desired photograph. Recently, Knorr and Kurz [KK14] proposed a framework for estimating the real-world lighting conditions based on the captured appearance of a human face. The method is based on learning a face-appearance model from an offline dataset of faces under known illumination.…”
Section: Estimated Lighting Conditionsmentioning
confidence: 99%
“…Lee and Woo [11] propose a method which reflects the illumination of a scene into the rendering of virtual objects by referring to the reflection on a fiducial marker. Knorr and Kurz [12] estimate a scene illumination from the reflection of light on human faces, and realize coherent AR visualization with the estimated illumination. Fischer et al [13] incorporate the sensor noise and the motion blur of a camera when rendering virtual objects by blurring and adding synthetic noise on an output image.…”
Section: Visual Consistency In Ar Applicationsmentioning
confidence: 99%