Learning Representations for Automatic Colorization
This work addresses the challenge of realistic colorization for images, which is an incremental improvement in computer vision.
The paper tackles the problem of automatic image colorization by developing a system that predicts per-pixel color histograms to handle multimodal color distributions, outperforming existing methods on fully and partially automatic tasks.
We develop a fully automatic image colorization system. Our approach leverages recent advances in deep networks, exploiting both low-level and semantic representations. As many scene elements naturally appear according to multimodal color distributions, we train our model to predict per-pixel color histograms. This intermediate output can be used to automatically generate a color image, or further manipulated prior to image formation. On both fully and partially automatic colorization tasks, we outperform existing methods. We also explore colorization as a vehicle for self-supervised visual representation learning.