Edge and identity preserving network for face super-resolution
Neurocomputing(2021)
摘要
•Edge block to compensate for high-frequency information.•Identity loss using JS divergence to preserve personal identities.•Luminance-chrominance error to separately learn brightness and color components.
更多查看译文
关键词
Super-resolution,Face hallucination,Edge block,Identity loss,Image enhancement
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要