US 12,079,306 B2
Methods and apparatuses of contrastive learning for color constancy
Yi-Chen Lo, Hsinchu (TW); and Chia-Che Chang, Hsinchu (TW)
Assigned to MEDIATEK INC., Hsinchu (TW)
Filed by MEDIATEK INC., Hsinchu (TW)
Filed on Nov. 18, 2021, as Appl. No. 17/529,539.
Claims priority of provisional application 63/116,377, filed on Nov. 20, 2020.
Prior Publication US 2022/0164601 A1, May 26, 2022
Int. Cl. G06T 7/90 (2017.01); G06F 18/214 (2023.01); G06F 18/22 (2023.01); G06N 20/00 (2019.01); G06V 10/40 (2022.01); G06V 10/56 (2022.01); H04N 9/73 (2023.01)
CPC G06F 18/214 (2023.01) [G06F 18/22 (2023.01); G06N 20/00 (2019.01); G06T 7/90 (2017.01); G06V 10/40 (2022.01); G06V 10/56 (2022.01); H04N 9/73 (2013.01); G06T 2207/10152 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 12 Claims
OG exemplary drawing
 
1. A contrastive learning method for color constancy in an image or video processing system, comprising:
receiving input data associated with a first training image captured in a first scene under a first illuminant, and a second training image captured in a second scene under a second illuminant;
constructing at least a positive contrastive pair and at least a negative contrastive pair by applying a data augmentation to the first and second training images, wherein each positive contrastive pair contains two images having an identical illuminant and each negative contrastive pair contains two images having different illuminants;
extracting representations of the images in the positive and negative contrastive pairs by a feature extraction function; and
training a color constancy model by contrastive learning, wherein the color constancy model is trained by learning representations in each positive contrastive pair are closer than the representations in each negative contrastive pair.