Joint Intensity-Gradient Guided Generative Modeling for Colorization
This work provides an incremental improvement in automatic image colorization for general users and applications by reducing artifacts and removing the need for reference images.
This paper introduces an iterative generative model for automatic colorization, addressing issues like edge color overflow and the need for reference images. By leveraging latent information in gradient maps and conducting inference in a joint intensity-gradient domain, the proposed system achieved superior performance compared to state-of-the-art methods in both quantitative comparisons and user studies.
This paper proposes an iterative generative model for solving the automatic colorization problem. Although previous researches have shown the capability to generate plausible color, the edge color overflow and the requirement of the reference images still exist. The starting point of the unsupervised learning in this study is the observation that the gradient map possesses latent information of the image. Therefore, the inference process of the generative modeling is conducted in joint intensity-gradient domain. Specifically, a set of intensity-gradient formed high-dimensional tensors, as the network input, are used to train a powerful noise conditional score network at the training phase. Furthermore, the joint intensity-gradient constraint in data-fidelity term is proposed to limit the degree of freedom within generative model at the iterative colorization stage, and it is conducive to edge-preserving. Extensive experiments demonstrated that the system outperformed state-of-the-art methods whether in quantitative comparisons or user study.