Exemplar-based image colorization refers to a computer-assisted process that adds colors to grayscale images. It is a challenging task since there is usually no one-to-one correspondence between color and local texture. Existing methods either perform poorly on some datasets or have strict requirements on reference images (content, position), which limits their practical application. In this talk, two methods are discussed to tackle such problems: superpixel-based variational model and style-based deep learning architecture. Extensive experiments are presented to demonstrate the effectiveness of the methods and to show these superiorities to other state-of-the-art methods.