TY - GEN
T1 - A Neural Network for Interpolating Light-Sources
AU - Colreavy-Donnelly, Simon
AU - Kuhn, Stefan
AU - Caraffini, Fabio
AU - Oconnor, Stuart
AU - Anastassi, Zacharias
AU - Coupland, Simon
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/7
Y1 - 2020/7
N2 - This study combines two novel deterministic methods with a Convolutional Neural Network to develop a machine learning method that is aware of directionality of light in images. The first method detects shadows in terrestrial images by using a sliding-window algorithm that extracts specific hue and value features in an image. The second method interpolates light-sources by utilising a line-algorithm, which detects the direction of light sources in the image. Both of these methods are single-image solutions and employ deterministic methods to calculate the values from the image alone, without the need for illumination-models. They extract real-time geometry from the light source in an image, rather than mapping an illuminationmodel onto the image, which are the only models used today. Finally, those outputs are used to train a Convolutional Neural Network. This displays greater accuracy than previous methods for shadow detection and can predict light source-direction and thus orientation accurately, which is a considerable innovation for an unsupervised CNN. It is significantly faster than the deterministic methods. We also present a reference dataset for the problem of shadow and light direction detection.
AB - This study combines two novel deterministic methods with a Convolutional Neural Network to develop a machine learning method that is aware of directionality of light in images. The first method detects shadows in terrestrial images by using a sliding-window algorithm that extracts specific hue and value features in an image. The second method interpolates light-sources by utilising a line-algorithm, which detects the direction of light sources in the image. Both of these methods are single-image solutions and employ deterministic methods to calculate the values from the image alone, without the need for illumination-models. They extract real-time geometry from the light source in an image, rather than mapping an illuminationmodel onto the image, which are the only models used today. Finally, those outputs are used to train a Convolutional Neural Network. This displays greater accuracy than previous methods for shadow detection and can predict light source-direction and thus orientation accurately, which is a considerable innovation for an unsupervised CNN. It is significantly faster than the deterministic methods. We also present a reference dataset for the problem of shadow and light direction detection.
KW - deep learning
KW - light source detection
KW - shadow detection
KW - single-image solution
KW - unsupervised learning
UR - http://www.scopus.com/inward/record.url?scp=85094097576&partnerID=8YFLogxK
U2 - 10.1109/COMPSAC48688.2020.00-21
DO - 10.1109/COMPSAC48688.2020.00-21
M3 - Conference contribution
AN - SCOPUS:85094097576
T3 - Proceedings - 2020 IEEE 44th Annual Computers, Software, and Applications Conference, COMPSAC 2020
SP - 1634
EP - 1640
BT - Proceedings - 2020 IEEE 44th Annual Computers, Software, and Applications Conference, COMPSAC 2020
A2 - Chan, W. K.
A2 - Claycomb, Bill
A2 - Takakura, Hiroki
A2 - Yang, Ji-Jiang
A2 - Teranishi, Yuuichi
A2 - Towey, Dave
A2 - Segura, Sergio
A2 - Shahriar, Hossain
A2 - Reisman, Sorel
A2 - Ahamed, Sheikh Iqbal
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 44th IEEE Annual Computers, Software, and Applications Conference, COMPSAC 2020
Y2 - 13 July 2020 through 17 July 2020
ER -