Anat Levin,Dani Lischinski和Yair Weiss在《Colorization using Optimization》一文中给出了一种基于人机交互的着色方案;
Alex Yong-Sang Chia,Shaojie Zhuo等人在《Semantic Colorization with Internet Images》中给出了一种利用网络彩色样本来给输入图片着色的方法;
Satoshi Iizuka, Edgar Simo-Serra和Hiroshi Ishikawa则是在《Let there be Color!: Joint End-to-end Learning of Global and Local Image Priors for Automatic Image Colorization with Simultaneous Classification》中提出了第一个能够自动着色且可进行端到端训练的模型;
L ( y c o l o r , y c l a s s ) = ∣ ∣ y c o l o r − y c o l o r , ∗ ∣ ∣ F R O 2 − α ( y l c l a s s c l a s s − log ( ∑ i = 0 N exp ( y i c l a s s ) ) ) \mathcal{L} (y^{\rm{color}}, y^{\rm{class}}) = ||y^{\rm{color}} - y^{\rm{color}, *} ||^{2}_{\rm{FRO}} - \alpha \left(y^{\rm{class}}_{l^{\rm{class}}} - \log \left(\sum^{N}_{i = 0} \exp (y^{\rm{class}}_{i})\right)\right) L(ycolor,yclass)=∣∣ycolor−ycolor,∗∣∣FRO2−α(ylclassclass−log(i=0∑Nexp(yiclass)))