Anycost GANs for Interactive Image Synthesis and Editing
Ji Lin, Richard Zhang, Frieder Ganz, Song Han, Jun-Yan Zhu
摘要
input projected 2x faster 4x faster 8x faster edit: no smile 8x faster edit: hair color 8x faster (a) input images (b) Anycost GANs provide consistent outputs for projected latent code (c) Consistency remains after editing less hair color Figure 1: Anycost GAN can be executed at flexible computation costs (fast preview with low cost and high-quality output with full cost), enabling interactive image editing with quick preview. The low-cost sub-generator produces consistent outputs compared to the full-cost generator during both image projection and latent code traversal, making the sub-generator an accurate proxy for various editing tasks (e.g., no smile, changing hair color). Interactive demos are available here.
问 Lune
几分钟读透这篇 Paper,不用再花一下午。
Lune 能读完全文,也能读引用它的顶会 Paper,讲清它的结论、证据和局限,每一点都注明出自哪篇 Paper。不用硬啃,也不靠猜。
已有 25 篇顶会 Paper 引用了它。别等审稿人问起,先弄清它们发现了什么。
脉络
- MCUNet: Tiny Deep Learning on IoT Devices 引用了 Once-for-All: Train One Network and Specialize it for Efficient Deployment
- Patch Diffusion: Faster and More Data-Efficient Training of Diffusion Models 引用了 Once-for-All: Train One Network and Specialize it for Efficient Deployment
- Patch Diffusion: Faster and More Data-Efficient Training of Diffusion Models 引用了 Training Generative Adversarial Networks with Limited Data
- HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image Editing 引用了 Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?
- HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image Editing 引用了 Training Generative Adversarial Networks with Limited Data
- The GAN is dead; long live the GAN! A Modern GAN Baseline 引用了 Training Generative Adversarial Networks with Limited Data
- Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models 引用了 MCUNet: Tiny Deep Learning on IoT Devices
- Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models 引用了 Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?
引用它的顶会 Paper25
- HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image EditingYuval Alaluf, Omer Tov, Ron Mokady, Rinon Gal 等CVPR 2022 · 250 次引用
- Patch Diffusion: Faster and More Data-Efficient Training of Diffusion ModelsZhendong Wang, Yifan Jiang, Huangjie Zheng, Peihao Wang 等NeurIPS 2023 · 205 次引用
- The GAN is dead; long live the GAN! A Modern GAN BaselineNick Huang, Aaron Gokaslan, Volodymyr Kuleshov, James TompkinNeurIPS 2024 · 111 次引用
- Adv-Attribute: Inconspicuous and Transferable Adversarial Attack on Face RecognitionShuai Jia, Bangjie Yin, Taiping Yao, Shouhong Ding 等NeurIPS 2022 · 84 次引用
- Efficient Spatially Sparse Inference for Conditional GANs and Diffusion ModelsMuyang Li, Ji Lin, Chenlin Meng, Stefano Ermon 等NeurIPS 2022 · 66 次引用
它引用的 Paper19
- Training Generative Adversarial Networks with Limited DataTero Karras, Miika Aittala, Janne Hellsten, Samuli Laine 等NeurIPS 2020 · 2,345 次引用
- Once-for-All: Train One Network and Specialize it for Efficient DeploymentHan Cai, Chuang Gan, Tianzhe Wang, Zhekai Zhang 等ICLR 2020 · 1,522 次引用
- Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?Rameen Abdal, Yipeng Qin, Peter WonkaICCV 2019 · 1,195 次引用
- GANSpace: Discovering Interpretable GAN ControlsErik Härkönen, Aaron Hertzmann, Jaakko Lehtinen, Sylvain ParisNeurIPS 2020 · 1,049 次引用
- MCUNet: Tiny Deep Learning on IoT DevicesJi Lin, Wei-Ming Chen, Yujun Lin, John Cohn 等NeurIPS 2020 · 827 次引用
相关 Paper
- Surrogate Gradient Field for Latent Space ManipulationMinjun Li, Yanghua Jin, Huachun ZhuCVPR 2021
- Using latent space regression to analyze and leverage compositionality in GANsLucy Chai, Jonas Wulff, Phillip IsolaICLR 2021 · 30 次引用
- GlassesGAN: Eyewear Personalization Using Synthetic Appearance Discovery and Targeted Subspace ModelingRichard Plesh, Peter Peer, Vitomir StrucCVPR 2023
- Auto DragGAN: Editing the Generative Image Manifold in an Autoregressive MannerPengxiang Cai, Zhiwei Liu, Guibo Zhu, Yunfang Niu 等ACM MM 2024 · 3 次引用
- Where is My Spot? Few-shot Image Generation via Latent Subspace OptimizationChenxi Zheng, Bangzhen Liu, Huaidong Zhang, Xuemiao Xu 等CVPR 2023
