Research map / Image generation / Fast diffusion sampling
CaricHarmony: Contrastive Diffusion Paths for Identity-Preserving Caricature Synthesis
Dongyu Wang, Dar-Yen Chen, Yi-Zhe Song
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 shape energy E_shape (ablation) | CaricHarmony · WebCaricature | S-CLIP: 0.8296 -> 0.8615 | Improved |
| add shape energy E_shape (ablation) | CaricHarmony · WebCaricature | I-CLIP: 0.7747 -> 0.7512 | Hurt |
| add shape energy E_shape (ablation) | CaricHarmony · WebCaricature | ImageReward: 0.8967 -> 0.8509 | Hurt |
| add ID energy E_id (ablation) | CaricHarmony · WebCaricature | I-CLIP: 0.7381 -> 0.7512 | Improved |
| add ID energy E_id (ablation) | CaricHarmony · WebCaricature | S-CLIP: 0.8698 -> 0.8615 | Hurt |
| add ID energy E_id (ablation) | CaricHarmony · WebCaricature | PickScore: 0.2199 -> 0.2049 | Hurt |
| add timestep-constrained guidance (TCG) (ablation) | CaricHarmony · WebCaricature | image quality | Improved |
| increase E_id scale while decreasing E_shape scale (ablation) | CaricHarmony · qualitative guidance scales | ID vs shape balance | Mixed |
| use CaricHarmony instead of DemoCaricature | CaricHarmony · WebCaricature | S-CLIP: 0.8450 -> 0.8615 | Improved |
| use CaricHarmony instead of DemoCaricature | CaricHarmony · WebCaricature | I-CLIP: 0.7591 -> 0.7512 | Hurt |
| use CaricHarmony instead of DemoCaricature | CaricHarmony · WebCaricature | ImageReward: 0.2871 -> 0.8509 | Improved |
| use CaricHarmony instead of DemoCaricature | CaricHarmony · User study | Shape score: 5.51 -> 8.08 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- DemoCaricature — DemoCaricature: Democratising Caricature Generation with a Rough Sketch (CVPR 2024) · better on 7 of 8 shared results
- WarpGAN — WarpGAN: Automatic Caricature Generation (CVPR 2019) · better on 3 of 3 shared results
- CaricatureBooth — CaricatureBooth: Data-Free Interactive Caricature Generation in a Photo Booth (CVPR 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DemoCaricature: Democratising Caricature Generation with a Rough Sketch (CVPR 2024 · similarity 0.87)
- CaricatureBooth: Data-Free Interactive Caricature Generation in a Photo Booth (CVPR 2025 · similarity 0.87)
- It's All About Your Sketch: Democratising Sketch Control in Diffusion Models (CVPR 2024 · similarity 0.87)
- DiffPortrait3D: Controllable Diffusion for Zero-Shot Portrait View Synthesis (CVPR 2024 · similarity 0.86)
- One-Shot Reference-based Structure-Aware Image to Sketch Synthesis (AAAI 2025 · similarity 0.86)
- StableAnimator: High-Quality Identity-Preserving Human Image Animation (CVPR 2025 · 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.
