Taming Adversarial Domain Transfer with Structural Constraints for Image Enhancement
This addresses image enhancement for traffic scenes, but the abstract only describes the problem and proposed methods without showing results, making it incremental.
This work tackles the problem of enhancing traffic scene images degraded by natural causes like fog, rain, and low visibility, where pixel-perfect paired data is unavailable, by using domain transfer models to transform degraded images to clear ones, but it does not report concrete results or numbers.
The goal of this work is to improve images of traffic scenes that are degraded by natural causes such as fog, rain and limited visibility during the night. For these applications, it is next to impossible to get pixel perfect pairs of the same scene, with and without the degrading conditions. This makes it unsuitable for conventional supervised learning approaches, however, it is easy to collect unpaired images of the scenes in a perfect and in a degraded condition. To enhance the images taken in a poor visibility condition, domain transfer models can be trained to transform an image from the degraded to the clear domain. A well-known concept for unsupervised domain transfer are cycle-consistent generative adversarial models. Unfortunately, the resulting generators often change the structure of the scene. This causes an undesirable change in the semantics. We propose three ways to cope with this problem depending on the type of degradation.