Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

WeditGAN: Few-Shot Image Generation via Latent Space Relocation (2305.06671v3)

Published 11 May 2023 in cs.CV

Abstract: In few-shot image generation, directly training GAN models on just a handful of images faces the risk of overfitting. A popular solution is to transfer the models pretrained on large source domains to small target ones. In this work, we introduce WeditGAN, which realizes model transfer by editing the intermediate latent codes $w$ in StyleGANs with learned constant offsets ($\Delta w$), discovering and constructing target latent spaces via simply relocating the distribution of source latent spaces. The established one-to-one mapping between latent spaces can naturally prevents mode collapse and overfitting. Besides, we also propose variants of WeditGAN to further enhance the relocation process by regularizing the direction or finetuning the intensity of $\Delta w$. Experiments on a collection of widely used source/target datasets manifest the capability of WeditGAN in generating realistic and diverse images, which is simple yet highly effective in the research area of few-shot image generation. Codes are available at https://github.com/Ldhlwh/WeditGAN.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (57)
  1. Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space? In ICCV.
  2. Image2StyleGAN++: How to Edit the Embedded Images? In CVPR.
  3. Data Augmentation Generative Adversarial Networks. arXiv preprint arXiv:1711.04340.
  4. Demystifying MMD GANs. In ICLR.
  5. Finding Directions in GAN’s Latent Space for Neural Face Reenactment. In BMVC.
  6. Image-Based CLIP-Guided Essence Transfer. In ECCV.
  7. JoJoGAN: One Shot Face Stylization. In ECCV.
  8. Editing in Style: Uncovering the Local Semantics of GANs. In CVPR.
  9. Attribute Group Editing for Reliable Few-shot Image Generation. In CVPR.
  10. Few-Shot Defect Image Generation via Defect-Aware Feature Manipulation. In AAAI.
  11. StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators. arXiv preprint arXiv:2108.00946.
  12. GANalyze: Toward Visual Definitions of Cognitive Image Properties. In ICCV.
  13. Generative Adversarial Nets. In NeurIPS.
  14. LoFGAN: Fusing Local Representations for Few-Shot Image Generation. In ICCV.
  15. GANSpace: Discovering Interpretable GAN Controls. In NeurIPS.
  16. GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium. In NeurIPS.
  17. Deltagan: Towards diverse few-shot image generation with sample-specific delta. In ECCV.
  18. MatchingGAN: Matching-based Few-shot Image Generation. In ICME.
  19. F2GAN: Fusing-and-Filling GAN for Few-Shot Image Generation. In ACM MM.
  20. Dynamic Weighted Semantic Correspondence for Few-Shot Image Generative Adaptation. In ACM MM.
  21. Training Generative Adversarial Networks with Limited Data. In NeurIPS.
  22. A Style-Based Generator Architecture for Generative Adversarial Networks. In CVPR.
  23. Analyzing and Improving the Image Quality of StyleGAN. In CVPR.
  24. One-Shot Adaptation of GAN in Just One CLIP. arXiv preprint arXiv:2203.09301.
  25. Few-shot Image Generation with Elastic Weight Consolidation. In NeurIPS.
  26. A Comprehensive Survey on Data-Efficient GANs in Image Generation. arXiv preprint arXiv:2204.08329.
  27. Style Intervention: How to Achieve Spatial Disentanglement with Style-based Generators? arXiv preprint arXiv:2011.09699.
  28. Deep Learning Face Attributes in the Wild. In ICCV.
  29. A Tutorial on Fisher information. Journal of Mathematical Psychology, 80: 40–55.
  30. Freeze the Discriminator: a Simple Baseline for Fine-Tuning GANs. In CVPR Workshop.
  31. Image Generation From Small Datasets via Batch Statistics Adaptation. In ICCV.
  32. Few-Shot Image Generation via Cross-Domain Correspondence. In CVPR.
  33. Multiple Heads are Better than One: Few-shot Font Generation with Multiple Localized Experts. In ICCV.
  34. StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery. In ICCV.
  35. Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation. In CVPR.
  36. Few-Shot Adaptation of Generative Adversarial Networks. arXiv preprint arXiv:2010.11943.
  37. Interpreting the Latent Space of GANs for Semantic Face Editing. In CVPR.
  38. Few-Shot Font Generation by Learning Fine-Grained Local Styles. In CVPR.
  39. StyleRig: Rigging StyleGAN for 3D Control over Portrait Images. In CVPR.
  40. Designing an Encoder for StyleGAN Image Manipulation. arXiv preprint arXiv:2102.02766.
  41. Face Photo-Sketch Synthesis and Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 31(11): 1955–1967.
  42. MineGAN: Effective Knowledge Transfer From GANs to Target Domains With Few Images. In CVPR.
  43. Transferring GANs: generating images from limited data. In ECCV.
  44. StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation. In CVPR.
  45. Few Shot Generative Model Adaption via Relaxed Spatial Structural Alignment. In CVPR.
  46. Pastiche Master: Exemplar-Based High-Resolution Portrait Style Transfer. In CVPR.
  47. The Face of Art: Landmark Detection and Geometric Style in Portraits. ACM Transactions on Graphics, 38(4).
  48. LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop. arXiv preprint arXiv:1506.03365.
  49. The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. In CVPR.
  50. Towards Diverse and Faithful One-shot Adaption of Generative Adversarial Networks. In NeurIPS.
  51. Generalized One-shot Domain Adaptation of Generative Adversarial Networks. In NeurIPS.
  52. On Leveraging Pretrained GANs for Generation with Limited Data. In ICML.
  53. Differentiable Augmentation for Data-Efficient GAN Training. In NeurIPS.
  54. Few-shot Image Generation via Adaptation-Aware Kernel Modulation. In NeurIPS.
  55. A Closer Look at Few-Shot Image Generation. In CVPR.
  56. Exploring Incompatible Knowledge Transfer in Few-shot Image Generation. In CVPR.
  57. Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for Generative Adversarial Networks. In ICLR.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (4)
  1. Yuxuan Duan (9 papers)
  2. Li Niu (79 papers)
  3. Yan Hong (49 papers)
  4. Liqing Zhang (80 papers)
Citations (9)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now