Research map / Video / Video generation
Durian: Dual Reference Image-Guided Portrait Animation with Attribute Transfer
Hyunsoo Cha, Byungjun Kim, Hanbyul Joo
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 |
|---|---|---|---|
| use Dual ReferenceNet instead of a single shared ReferenceNet (ablation) | Durian · Self-attribute transfer | PSNR: 17.95 -> 18.83 | Improved |
| add attribute-aware mask expansion (ablation) | Durian · Self-attribute transfer | PSNR: 17.16 -> 18.83 | Improved |
| add reference image augmentation (ablation) | Durian · Self-attribute transfer | PSNR: 16.97 -> 18.83 | Improved |
| add binary mask input to ReferenceNets (ablation) | Durian · Self-attribute transfer | PSNR: 18.60 -> 18.83 | Improved |
| use complementary masked references instead of full unmasked reference images (ablation) | Durian · Self-attribute transfer | PSNR: 19.47 -> 18.83 | Mixed |
| Durian single-stage vs best two-stage baseline on cross-attribute transfer | portrait animation with attribute transfer · Cross-attribute transfer | mDINO: 0.7271 -> 0.7801 | Improved |
| Durian vs two-stage baselines on temporal realism (VFID ResNeXt) | portrait animation with attribute transfer · Cross-attribute transfer | VFID (ResNeXt): 1.4672 (StableHair+MegActor) vs 2.4052 | Hurt |
Closest papers by meaning
- EchoShot: Multi-Shot Portrait Video Generation (NeurIPS 2025 · similarity 0.85)
- High-Fidelity Relightable Monocular Portrait Animation with Lighting-Controllable Video Diffusion Model (CVPR 2025 · similarity 0.85)
- RealPortrait: Realistic Portrait Animation with Diffusion Transformers (AAAI 2025 · similarity 0.85)
- X-NeMo: Expressive Neural Motion Reenactment via Disentangled Latent Attention (ICLR 2025 · similarity 0.85)
- ARGUS: Stacked Multi-View Identity Mosaic Injection for Subject-Preserving Video Generation (NeurIPS 2026 · similarity 0.84)
- Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer (CVPR 2025 · similarity 0.84)
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.
