Research map / Label-efficient and robust learning / Recognition and long-tail
EmbryoDiff: A Conditional Diffusion Framework with Multi-Focal Feature Fusion for Fine-Grained Embryo Developmental Stage Recognition
Yong Sun, Zhengjie Zhang, Junyu Shi, Zhiyuan Zhang, Lijiang Liu, Qiang Nie
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 multi-focal feature fusion to the semantic-encoder-only model (ablation) | SCE-only model · Multi-Focal Human Embryos (MFHE), 15 steps | accuracy: 80.1 -> 81.3 | Improved |
| add multi-focal feature fusion to the diffusion model (ablation) | EmbryoDiff · Multi-Focal Human Embryos (MFHE), 15 steps | accuracy: 81.0 -> 82.5 | Improved |
| add conditional diffusion model over semantic encoder only (ablation) | SCE-only model · Multi-Focal Human Embryos (MFHE), 15 steps | accuracy: 80.1 -> 81.0 | Improved |
| add the boundary condition encoder to the diffusion model without MFF (ablation) | EmbryoDiff · Multi-Focal Human Embryos (MFHE), 15 steps | Edit score: 82.8 -> 83.3 | Improved |
| add multi-focal fusion to the full model with boundary encoder (ablation) | EmbryoDiff · Multi-Focal Human Embryos (MFHE), 15 steps | accuracy: 80.6 -> 83.1 | Improved |
| add multi-focal fusion to the full model with boundary encoder (ablation) | EmbryoDiff · Multi-Focal Human Embryos (MFHE), 15 steps | F1@50: 63.9 -> 68.4 | Improved |
| add conditional diffusion model on single-focal data (ablation) | SCE-only model · Single-Focal Human Embryos (SFHE), 15 steps | Edit score: 81.8 -> 85.8 | Improved |
| add boundary condition encoder on single-focal data (ablation) | EmbryoDiff · Single-Focal Human Embryos (SFHE), 15 steps | F1@50: 70.2 -> 71.2 | Improved |
| increase fused focal planes from 1 to 7 (ablation) | EmbryoDiff · Multi-Focal Human Embryos (MFHE), 15 steps | accuracy: 80.6 -> 83.1 | Improved |
| increase fused focal planes from 1 to 7 (ablation) | EmbryoDiff · Multi-Focal Human Embryos (MFHE), 15 steps | Edit score: 83.3 -> 86.3 | Improved |
| increase inference denoising steps from 1 to 25 | EmbryoDiff · Multi-Focal Human Embryos (MFHE) | average metric: 77.7 -> 80.2 | Improved |
| increase inference denoising steps beyond about 20 | EmbryoDiff · Multi-Focal Human Embryos (MFHE) | accuracy: 82.8 -> 83.1 | No clear effect |
Closest papers by meaning
- Semantic and Visual Crop-Guided Diffusion Models for Heterogeneous Tissue Synthesis in Histopathology (NeurIPS 2025 · similarity 0.84)
- MorphoHELM: A Comprehensive Benchmark for Evaluating Representations for Microscopy-Based Morphology Assays (NeurIPS 2026 · similarity 0.83)
- Multimodal Prototyping for cancer survival prediction (ICML 2024 · similarity 0.82)
- Modeling Dense Multimodal Interactions Between Biological Pathways and Histology for Survival Prediction (CVPR 2024 · similarity 0.82)
- F$^2$-Assist: Multi-Phase Fetal Growth Forecast and Report Generation from Ultrasound Examination (CVPR 2026 · similarity 0.82)
- ViTally Consistent: Scaling Biological Representation Learning for Cell Microscopy (ICML 2025 · similarity 0.81)
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.
