Research map / Detection and segmentation / CNN features
Instance Segmentation of Biological Images Using Harmonic Embeddings
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 optimized harmonic guide embeddings with SinConv | U-Net instance segmentation on CVPPP A1 · CVPPP A1 | SBD (%): 84.9 -> 89.9 | Helped |
| Use harmonic embeddings instead of the U-Net baseline | Instance segmentation on E. coli · E. coli | SBD (%): 59.3 -> 81.2 | Helped |
| Use harmonic embeddings instead of Mask R-CNN | Instance segmentation on C. elegans · C. elegans | AP: 0.559 -> 0.724 | Helped |
| Apply a foreground mask to the predicted segmentation | Harmonic Embeddings on HeLa · HeLa | SBD (%): 78% -> 86% | Helped |
| Use harmonic embeddings instead of Mask R-CNN | Instance segmentation on COCO val person class · COCO val (person) | AP: 0.319 -> 0.223 | Hurt |
| Add SinConv guide-function conditioning rather than using no guide (ablation) | Harmonic Embeddings on CVPPP A1 · CVPPP A1 | SBD (%): 78.3 -> 89.9 | Helped |
| Use SinConv instead of CoordConv (ablation) | Harmonic Embeddings on CVPPP A1 · CVPPP A1 | SBD (%): 87.6 -> 89.9 | Helped |
| Optimize guide functions instead of using random guide functions (ablation) | Harmonic Embeddings on CVPPP A1 · CVPPP A1 | SBD (%): 70.8 -> 89.9 | Helped |
| Use optimized guides instead of low-frequency guide functions (ablation) | Harmonic Embeddings on CVPPP A1 · CVPPP A1 | SBD (%): 86.2 -> 89.9 | Helped |
| Use optimized guides instead of high-frequency guide functions (ablation) | Harmonic Embeddings on CVPPP A1 · CVPPP A1 | SBD (%): 5.4 -> 89.9 | Helped |
| Add SinConv guide-function conditioning rather than using no guide (ablation) | Harmonic Embeddings on C. elegans · C. elegans | AP: 0.028 -> 0.724 | Helped |
| Use SinConv instead of CoordConv (ablation) | Harmonic Embeddings on C. elegans · C. elegans | AP: 0.436 -> 0.724 | Helped |
Closest papers by meaning
- Object-Guided Instance Segmentation for Biological Images (AAAI 2020 · similarity 0.90)
- Sparse Object-Level Supervision for Instance Segmentation With Pixel Embeddings (CVPR 2022 · similarity 0.87)
- Learning to Model Pixel-Embedded Affinity for Homogeneous Instance Segmentation (AAAI 2022 · similarity 0.87)
- The Four Color Theorem for Cell Instance Segmentation (ICML 2025 · similarity 0.85)
- DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Data (CVPR 2024 · similarity 0.82)
- $\epsilon$-Seg: Sparsely Supervised Semantic Segmentation of Microscopy Data (NeurIPS 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.