Research map / Detection and segmentation / CNN features
Instance Segmentation by Jointly Optimizing Spatial Embeddings and Clustering Bandwidth
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 an instance-specific learnable sigma instead of a fixed sigma (ablation) | Spatial embedding instance segmentation · Cityscapes validation, ground-truth sampling | AP_gt: 28.0 -> 38.7 | Helped |
| Use a learnable center of attraction instead of the centroid with scalar sigma (ablation) | Spatial embedding instance segmentation · Cityscapes validation, ground-truth sampling | AP_gt: 38.7 -> 39.5 | Helped |
| Use a learnable center of attraction instead of the centroid with 2D sigma (ablation) | Spatial embedding instance segmentation · Cityscapes validation, ground-truth sampling | AP_gt: 39.1 -> 40.5 | Helped |
| Use 2D sigma instead of scalar sigma with centroid center (ablation) | Spatial embedding instance segmentation · Cityscapes validation, ground-truth sampling | AP_gt: 38.7 -> 39.1 | Helped |
| Use 2D sigma instead of scalar sigma with learnable center (ablation) | Spatial embedding instance segmentation · Cityscapes validation, ground-truth sampling | AP_gt: 39.5 -> 40.5 | Helped |
| Use the proposed method rather than Mask R-CNN trained on fine annotations | Cityscapes instance segmentation · Cityscapes test | AP: 26.2 -> 27.6 | Helped |
| Use the proposed method rather than PANet trained on fine annotations | Cityscapes instance segmentation · Cityscapes test | AP: 31.8 -> 27.6 | Hurt |
Closest papers by meaning
- Instance Segmentation With Mask-Supervised Polygonal Boundary Transformers (CVPR 2022 · similarity 0.89)
- Pixelwise Instance Segmentation With a Dynamically Instantiated Network (CVPR 2017 · similarity 0.88)
- SOLOv2: Dynamic and Fast Instance Segmentation (NeurIPS 2020 · similarity 0.88)
- Joint 3D Instance Segmentation and Object Detection for Autonomous Driving (CVPR 2020 · similarity 0.88)
- Training-Free Dataset Pruning for Instance Segmentation (ICLR 2025 · similarity 0.86)
- Any3DIS: Class-Agnostic 3D Instance Segmentation by 2D Mask Tracking (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.