Research map / Image generation / Autoregressive and latent generation
RDFace: A Benchmark Dataset for Rare Disease Facial Image Analysis under Extreme Data Scarcity and Phenotype-Aware Synthetic Generation
Ganlin Feng, Yuxi Long, Hafsa Ali, Erin Lou, Fahad Butt, Qian Liu, Yang Wang, Pingzhao Hu
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 DreamBooth synthetic images (Top-1000) to training (ablation) | DenseNet-169 · RDFace (supervised) | Top-1 accuracy (%): 15.93 -> 17.52 | Improved |
| add DreamBooth synthetic images (Top-1000) to training (ablation) | VGG-16 · RDFace (supervised) | Top-1 accuracy (%): 11.68 -> 16.64 | Improved |
| add FastGAN unconditional synthetic images to training (ablation) | DenseNet-169 · RDFace (supervised) | Top-1 accuracy (%): 15.93 -> 13.27 | Hurt |
| add both DreamBooth and FastGAN images (ablation) | DenseNet-169 · RDFace (supervised) | Top-1 accuracy (%): 15.93 -> 16.46 | Improved |
| use generic MixUp or CutMix augmentation (ablation) | DenseNet-169 · RDFace (supervised) | Top-1 accuracy (%): 15.93 -> 15.75 (MixUp), 16.11 (CutMix) | No clear effect |
| scale DreamBooth images from Top-1000 to Top-6000 (ablation) | DenseNet-169 · RDFace (supervised) | Top-1 accuracy (%): 17.52 -> 21.06 | Improved |
| add DreamBooth images to the support set (ablation) | DenseNet-169 ProtoNet · RDFace (5-way 1-shot) | accuracy (%): 26.20 -> 29.88 | Improved |
| add DreamBooth images to the support set (ablation) | FaceNet, VGG-16, CLIP ProtoNet · RDFace (5-way 1-shot) | accuracy (%): FaceNet 25.16 -> 23.60; VGG 21.54 -> 21.02; CLIP 23.48 -> 22.30 | Hurt |
| condition generation on class (DreamBooth) versus unconditional FastGAN (ablation) | synthetic augmentation · RDFace expert review | plausible rate: FastGAN 2-38% -> DreamBooth 62-76% | Improved |
Closest papers by meaning
- Doctor Approved: Generating Medically Accurate Skin Disease Images through AI-Expert Feedback (NeurIPS 2025 · similarity 0.84)
- BIOMEDICA: An Open Biomedical Image-Caption Archive, Dataset, and Vision-Language Models Derived from Scientific Literature (CVPR 2025 · similarity 0.84)
- DermaCon-IN: A Multiconcept-Annotated Dermatological Image Dataset of Indian Skin Disorders for Clinical AI Research (NeurIPS 2025 · similarity 0.83)
- Unraveling Normal Anatomy via Fluid-Driven Anomaly Randomization (CVPR 2025 · similarity 0.83)
- Breaking the Barrier of Hard Samples: A Data-Centric Approach to Synthetic Data for Medical Tasks (ICML 2025 · similarity 0.83)
- Medical Manifestation-Aware De-Identification (AAAI 2025 · similarity 0.82)
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.
