One of the essential image processing algorithms in every digital camera is the auto white balance (AWB) also known as color constancy. This algorithm makes sure that the colors in the image you captured look natural. In his PhD thesis Mr. Yanlin Qian (Msc in Technology) introduces two completely novel approaches: temporal color constancy and grayness index. Mr. Qian’s algorithms are state-of-the-art methods for color constancy and were thus published in the premier forums of computer vision and machine learning, IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) 2019 and Int. Conf. on Computer Vision (ICCV) 2017. Thanks to these remarkable results Mr. Qian was immediately after submitting his thesis recruited to the top camera lab of Huawei whose mobile phone cameras are number one in the world at the moment.
“Yes, I was lucky to invent these two ideas during my PhD thesis and to demonstrate to the scientific community how powerful they actually are!” He continues: “It is noteworthy that while temporal color constancy is AI-powered the Grayness Index is traditional based on physics of color reproduction. I didn’t want to be AI-only, but also wanted to understand the phenomenon itself and thus leading to two very different ideas but equally powerful!," Yanlin Qian replies from Beijing.
In his work Mr. Qian continues to collaborate with the imaging researchers of Tampere University and also the Huawei Camera team in Tampere.
The doctoral dissertation of MSc (Tech) Yanlin Qian in the field of Information Technology titled Computational Color Constancy: From Pixel to Video with a Stop at Convolutional Neural Network will be publicly examined in the Faculty of Information Technology and Computing Sciences at Tampere University at 11 December 2020 noon in public online defense. The Opponents will be Professor Ales Leonardis from University of Birmingham and Assistant Professor Juho Kannala from Aalto University. The Custos will be Professor Joni Kämäräinen.
The dissertation is available online at the http://urn.fi/URN:ISBN:978-952-03-1791-1
Photo: Yanlin Qian’s personal album