TY - GEN
T1 - A Generative Adversarial Framework for Optimizing Image Matting and Harmonization Simultaneously
AU - Ren, Xuqian
AU - Liu, Yifan
AU - Song, Chunlei
N1 - Extension for accepted ICIP 2021
PY - 2021/8/13
Y1 - 2021/8/13
N2 - Image matting and image harmonization are two important tasks in image composition. Image matting, aiming to achieve foreground boundary details, and image harmonization, aiming to make the background compatible with the foreground, are both promising yet challenging tasks. Previous works consider optimizing these two tasks separately, which may lead to a sub-optimal solution. We propose to optimize matting and harmonization simultaneously to get better performance on both the two tasks and achieve more natural results. We propose a new Generative Adversarial (GAN) framework which optimizing the matting network and the harmonization network based on a self-attention discriminator. The discriminator is required to distinguish the natural images from different types of fake synthesis images. Extensive experiments on our constructed dataset demonstrate the effectiveness of our proposed method. Our dataset and dataset generating pipeline can be found in \url{https://git.io/HaMaGAN}
AB - Image matting and image harmonization are two important tasks in image composition. Image matting, aiming to achieve foreground boundary details, and image harmonization, aiming to make the background compatible with the foreground, are both promising yet challenging tasks. Previous works consider optimizing these two tasks separately, which may lead to a sub-optimal solution. We propose to optimize matting and harmonization simultaneously to get better performance on both the two tasks and achieve more natural results. We propose a new Generative Adversarial (GAN) framework which optimizing the matting network and the harmonization network based on a self-attention discriminator. The discriminator is required to distinguish the natural images from different types of fake synthesis images. Extensive experiments on our constructed dataset demonstrate the effectiveness of our proposed method. Our dataset and dataset generating pipeline can be found in \url{https://git.io/HaMaGAN}
KW - cs.CV
U2 - 10.1109/ICIP42928.2021.9506642
DO - 10.1109/ICIP42928.2021.9506642
M3 - Conference contribution
T3 - ICIP
SP - 1354
EP - 1358
BT - 2021 IEEE International Conference on Image Processing (ICIP)
PB - IEEE
T2 - IEEE International Conference on Image Processing
Y2 - 19 September 2021 through 22 September 2021
ER -