INTERACTIVE SEPARATION NETWORK FOR IMAGE INPAINTING
Published in 27th IEEE International Conference on Image Processing, 2020
Recommended citation: Li, Siyuan, et al. INTERACTIVE SEPARATION NETWORK FOR IMAGE INPAINTING.[C], 27th IEEE International Conference on Image Processing (ICIP 2020),2020.
First author, Image inpainting paper in CCF C conference paper.
Temporary Citation:
Siyuan Li, Lu Lu, Zhiqiang Zhang, Xin Cheng, Kepeng Xu, Wenxin Yu, Gang He, Yibo Fan, Zhuo Yang. INTERACTIVE SEPARATION NETWORK FOR IMAGE INPAINTING.[C], 27th IEEE International Conference on Image Processing (ICIP 2020),2020. (CCF 推荐C类推荐会议)(第一作者身份)
Introduction:
被ICIP2020接受的学术论文中，在图像修复领域提出了新的深度学习图像修复生成器模型（交互分离网络），模型采用双流交替交互信息的正向传播方式，旨在时网络中的特征在具备高级语义信息的同时也能兼顾高分辨的低级纹理特征。模型的正向传播的每个阶段都加入了残差模块，并研究的加入数量对修复效果的具体影响。该方法提高了图像的修复质量和修复效果，并且论文中的方法在PSNR和SSIM两个量化指标上都高于已知的前沿修复技术。
Download paper here