Discovering Interpretable Latent Space Directions of GANs Beyond Binary Attributes
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021（2021）
关键词
controllable representation enabling manipulation,nonbinary attributes,anime styles,facial characteristics,multiattribute manipulation,interpretable latent space directions,GANs,generative adversarial networks,high-fidelity image outputs,input latent space,semantic correlations,output image space,human interpretable image transformations,editing tasks,free-form artistic tasks,adversarial method,noise latent vector mapping,style/anime editing
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