Deep Feature Consistent Deep Image Transformations: Downscaling, Decolorization and HDR Tone Mapping
This work addresses visual integrity issues in image processing for applications like graphics and photography, presenting a novel deep learning approach that unifies multiple tasks, though it is incremental in combining existing CNN insights.
The paper tackled the problem of unifying challenging one-to-many mapping image processing tasks, including image downscaling, decolorization, and HDR tone mapping, by developing the DFC-DIT framework, which achieved state-of-the-art performances as demonstrated in experimental results.
Building on crucial insights into the determining factors of the visual integrity of an image and the property of deep convolutional neural network (CNN), we have developed the Deep Feature Consistent Deep Image Transformation (DFC-DIT) framework which unifies challenging one-to-many mapping image processing problems such as image downscaling, decolorization (colour to grayscale conversion) and high dynamic range (HDR) image tone mapping. We train one CNN as a non-linear mapper to transform an input image to an output image following what we term the deep feature consistency principle which is enforced through another pretrained and fixed deep CNN. This is the first work that uses deep learning to solve and unify these three common image processing tasks. We present experimental results to demonstrate the effectiveness of the DFC-DIT technique and its state of the art performances.