TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Image Harmonization	HAdobe5k(1024$\times$1024)	DoveNet	PSNR	34.81	# 6
Image Harmonization	HAdobe5k(1024$\times$1024)	DoveNet	MSE	51.00	# 6
Image Harmonization	HAdobe5k(1024$\times$1024)	DoveNet	fMSE	312.88	# 2
Image Harmonization	HAdobe5k(1024$\times$1024)	DoveNet	SSIM	0.9729	# 6
Image Harmonization	iHarmony4	DoveNet	MSE	52.33	# 16
Image Harmonization	iHarmony4	DoveNet	PSNR	34.76	# 16
Image Harmonization	iHarmony4	DoveNet	fMSE	549.96	# 1

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/deep-image-harmonization-via-domain/image-harmonization-on-hadobe5k-1024-times)](https://paperswithcode.com/sota/image-harmonization-on-hadobe5k-1024-times?p=deep-image-harmonization-via-domain)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/deep-image-harmonization-via-domain/image-harmonization-on-iharmony4)](https://paperswithcode.com/sota/image-harmonization-on-iharmony4?p=deep-image-harmonization-via-domain)`

DoveNet: Deep Image Harmonization via Domain Verification

CVPR 2020 · Wenyan Cong, Jianfu Zhang, Li Niu, Liu Liu, Zhixin Ling, Weiyuan Li, Liqing Zhang ·

Image composition is an important operation in image processing, but the inconsistency between foreground and background significantly degrades the quality of composite image. Image harmonization, aiming to make the foreground compatible with the background, is a promising yet challenging task. However, the lack of high-quality publicly available dataset for image harmonization greatly hinders the development of image harmonization techniques. In this work, we contribute an image harmonization dataset iHarmony4 by generating synthesized composite images based on COCO (resp., Adobe5k, Flickr, day2night) dataset, leading to our HCOCO (resp., HAdobe5k, HFlickr, Hday2night) sub-dataset. Moreover, we propose a new deep image harmonization method DoveNet using a novel domain verification discriminator, with the insight that the foreground needs to be translated to the same domain as background. Extensive experiments on our constructed dataset demonstrate the effectiveness of our proposed method. Our dataset and code are available at https://github.com/bcmi/Image_Harmonization_Datasets.