Density-aware Haze Image Synthesis by Self-Supervised Content-Style Disentanglement

TL;DR

This paper introduces a self-supervised style regression method for haze image synthesis that improves disentanglement of content and style features, leading to more accurate haze rendering and better generalization of vehicle detectors under hazy conditions.

Contribution

It proposes a novel self-supervised style regression technique to enhance content-style disentanglement in haze image synthesis, addressing limitations of previous methods.

Findings

Improved haze image synthesis quality with complete content-style disentanglement.

Haze level significantly affects vehicle detection performance, with degradation linearly correlated to haze severity.

Generated haze data enhances the robustness of vehicle detectors in hazy environments.

Abstract

The key procedure of haze image translation through adversarial training lies in the disentanglement between the feature only involved in haze synthesis, i.e.style feature, and the feature representing the invariant semantic content, i.e. content feature. Previous methods separate content feature apart by utilizing it to classify haze image during the training process. However, in this paper we recognize the incompleteness of the content-style disentanglement in such technical routine. The flawed style feature entangled with content information inevitably leads the ill-rendering of the haze images. To address, we propose a self-supervised style regression via stochastic linear interpolation to reduce the content information in style feature. The ablative experiments demonstrate the disentangling completeness and its superiority in level-aware haze image synthesis. Moreover, the generated haze data are applied in the testing generalization of vehicle detectors. Further study between haze-level and detection performance shows that haze has obvious impact on the generalization of the vehicle detectors and such performance degrading level is linearly correlated to the haze-level, which, in turn, validates the effectiveness of the proposed method.