Research map / Label-efficient and robust learning / Domain adaptation and generalization
Few Shot Generative Model Adaption via Relaxed Spatial Structural Alignment
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use RSSA instead of IDC | StyleGANv2 few-shot adaptation · Faces → Sketches, 10-shot | SCS: 0.437 -> 0.511 | Helped |
| Use RSSA instead of IDC | StyleGANv2 few-shot adaptation · Churches → Haunted Houses, 5-shot | SCS: 0.424 -> 0.582 | Helped |
| Use RSSA instead of the strongest baseline | StyleGANv2 few-shot adaptation · Faces → Van Gogh, 10-shot | IS: 2.69$^{\pm0.18}$ -> 1.77$^{\pm0.06}$ | Hurt |
| Remove spatial structural consistency loss (ablation) | RSSA · Faces → Van Gogh, 10-shot | SCS: 0.449 -> 0.702 | Hurt |
| Add self-correlation consistency loss (ablation) | RSSA without spatial structural consistency loss · Faces → Van Gogh, 10-shot | SCS: 0.449 -> 0.684 | Helped |
| Add disturbance-correlation consistency loss (ablation) | RSSA without spatial structural consistency loss · Faces → Van Gogh, 10-shot | SCS: 0.449 -> 0.671 | Helped |
| Use both structural consistency losses instead of self-correlation loss only (ablation) | RSSA · Faces → Van Gogh, 10-shot | SCS: 0.684 -> 0.702 | Helped |
| Use both structural consistency losses instead of disturbance-correlation loss only (ablation) | RSSA · Churches → Haunted Houses, 10-shot | SCS: 0.548 -> 0.573 | Helped |
| Add latent space compression (ablation) | RSSA · Faces → Sketches, 10-shot | adaptation speed | Helped |
Closest papers by meaning
- Few-Shot Image Generation via Cross-Domain Correspondence (CVPR 2021 · similarity 0.89)
- Progressive Few-Shot Adaptation of Generative Model with Align-Free Spatial Correlation (AAAI 2023 · similarity 0.89)
- Few-shot Image Generation via Adaptation-Aware Kernel Modulation (NeurIPS 2022 · similarity 0.89)
- Towards Diverse and Faithful One-shot Adaption of Generative Adversarial Networks (NeurIPS 2022 · similarity 0.89)
- WeditGAN: Few-Shot Image Generation via Latent Space Relocation (AAAI 2024 · similarity 0.87)
- Aligning Few-Step Generative Model via Amortizing Sample-Based Variational Inference (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.