Deep Image Harmonization in Dual Color Spaces

Abstract

Image harmonization is an essential step in image composition that adjusts the appearance of composite foreground to address the inconsistency between foreground and background. Existing methods primarily operate in correlated $RGB$ color space, leading to entangled features and limited representation ability. In contrast, decorrelated color space (e.g., $Lab$) has decorrelated channels that provide disentangled color and illumination statistics. In this paper, we explore image harmonization in dual color spaces, which supplements entangled $RGB$ features with disentangled $L$, $a$, $b$ features to alleviate the workload in harmonization process. The network comprises a $RGB$ harmonization backbone, an $Lab$ encoding module, and an $Lab$ control module. The backbone is a U-Net network translating composite image to harmonized image. Three encoders in $Lab$ encoding module extract three control codes independently from $L$, $a$, $b$ channels, which are used to manipulate the decoder features in harmonization backbone via $Lab$ control module. Our code and model are available at \href{https://github.com/bcmi/DucoNet-Image-Harmonization}{https://github.com/bcmi/DucoNet-Image-Harmonization}.

arXiv commentAccepted by ACMMM 2023

References in corpus (9)

• Underwater Image Enhancement via Medium Transmission-Guided Multi-Color Space Embedding
• U-shape Transformer for Underwater Image Enhancement
• Adaptive Computation Time for Recurrent Neural Networks
• Improving the Harmony of the Composite Image by Spatial-Separated Attention Module
• Making Images Real Again: A Comprehensive Survey on Deep Image Composition
• Dense Pixel-to-Pixel Harmonization via Continuous Image Representation
• Image Harmonization by Matching Regional References
• FRIH: Fine-grained Region-aware Image Harmonization
• Inharmonious Region Localization with Auxiliary Style Feature