Research map / Image generation / Text-to-image diffusion
UniCSG: Unified High-Fidelity content-constrained style-driven generation via Staged Semantic and Frequency Disentanglement
Jingwei Yang, Ruoxi Wu, Wei Shen, Meng Li, Yulong Liu, Huimin She, Lunxi Yuan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add Stage 1 semantic disentanglement (ablation) | UniCSG · CSG-Bench (text-guided) | DINO: 0.590 -> 0.701 | Improved |
| add Stage 1 semantic disentanglement (ablation) | UniCSG · CSG-Bench (text-guided) | CLIP-I: 0.748 -> 0.797 | Improved |
| add Stage 2 frequency-aware reconstruction (ablation) | UniCSG · CSG-Bench (text-guided) | FID: 124.718 -> 113.940 | Improved |
| add Stage 2 frequency-aware reconstruction (ablation) | UniCSG · CSG-Bench (text-guided) | CSD: 0.516 -> 0.541 | Improved |
| add pixel-space reward learning (ablation) | UniCSG · CSG-Bench (text-guided) | FID: 123.331 -> 113.940 | Improved |
| add pixel-space reward learning (ablation) | UniCSG · CSG-Bench (text-guided) | DreamSim: 0.775 -> 0.816 | Improved |
| add Stage 1 semantic disentanglement (ablation) | UniCSG · CSG-Bench (reference-guided) | DINO: 0.417 -> 0.597 | Improved |
| add Stage 1 semantic disentanglement (ablation) | UniCSG · CSG-Bench (reference-guided) | CLIP-I: 0.716 -> 0.760 | Improved |
| add Stage 2 frequency-aware reconstruction (ablation) | UniCSG · CSG-Bench (reference-guided) | CSD: 0.619 -> 0.731 | Improved |
| add Stage 2 frequency-aware reconstruction (ablation) | UniCSG · CSG-Bench (reference-guided) | DINO: 0.452 -> 0.597 | Improved |
| add pixel-space reward learning (ablation) | UniCSG · CSG-Bench (reference-guided) | FID: 92.335 -> 87.320 | Improved |
| add pixel-space reward learning (ablation) | UniCSG · CSG-Bench (reference-guided) | CSD: 0.661 -> 0.731 | Improved |
Closest papers by meaning
- Conditional Balance: Improving Multi-Conditioning Trade-Offs in Image Generation (CVPR 2025 · similarity 0.88)
- SigStyle: Signature Style Transfer via Personalized Text-to-Image Models (AAAI 2025 · similarity 0.88)
- CLeaR: A Unified Framework for Resolving the Leakage–Degradation Dilemma in Style Transfer (NeurIPS 2026 · similarity 0.87)
- CSGO: Content-Style Composition in Text-to-Image Generation (NeurIPS 2025 · similarity 0.87)
- Unsupervised Disentanglement of Content and Style via Variance-Invariance Constraints (ICLR 2025 · similarity 0.87)
- Disentangling Dual Image References in Frequency Aware Diffusion Models for Personalized Generation (NeurIPS 2026 · similarity 0.87)
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.
