Research map / Detection and segmentation / CNN features
PointFlow: Flowing Semantics Through Points for Aerial Image Segmentation
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 |
|---|---|---|---|
| Add PFNet point-flow propagation to SemanticFPN | SemanticFPN, iSAID val · iSAID val | mIoU: 61.3 -> 65.0 | Helped |
| Replace dense affinity with dense affinity propagation | SemanticFPN, iSAID val · iSAID val | mIoU: 61.3 -> 58.9 | Hurt |
| Add both salient and boundary point flows (ablation) | SemanticFPN + PPM, iSAID val · iSAID val | mIoU: 63.8 -> 66.9 | Helped |
| Add boundary point flow (ablation) | SemanticFPN + PPM, iSAID val · iSAID val | mIoU: 64.8 -> 66.9 | Helped |
| Add salient point flow (ablation) | SemanticFPN + PPM, iSAID val · iSAID val | mIoU: 66.2 -> 66.9 | Helped |
| Add PPM context head (ablation) | SemanticFPN + point flows, iSAID val · iSAID val | mIoU: 65.0 -> 66.9 | Helped |
| Insert PFM at all tested pyramid stages rather than only one stage (ablation) | SemanticFPN + PPM, iSAID val · iSAID val | mIoU: 65.8 -> 66.9 | Helped |
| Use max pooling for salient-point selection instead of random or attention-based sampling (ablation) | PFM, iSAID val · iSAID val | mIoU: 64.0 -> 64.8; 64.2 -> 64.8 | Helped |
| Propagate semantics top-down rather than bottom-up (ablation) | PFM, iSAID val · iSAID val | mIoU: 47.3 -> 66.9 | Helped |
| Use subtraction-based boundary prediction instead of direct prediction (ablation) | PFM, iSAID val · iSAID val | mIoU: 65.7 -> 66.2 | Helped |
| Use PFM rather than dense affinity flow (ablation) | SemanticFPN + PPM, iSAID val · iSAID val | mIoU: 62.0 -> 66.9 | Helped |
| Add PF decoder to Deeplabv3 | Deeplabv3, ResNet50, iSAID val · iSAID val | mIoU: 60.4 -> 65.6 | Helped |
Closest papers by meaning
- Aerial Lifting: Neural Urban Semantic and Building Instance Lifting from Aerial Imagery (CVPR 2024 · similarity 0.86)
- SatSynth: Augmenting Image-Mask Pairs through Diffusion Models for Aerial Semantic Segmentation (CVPR 2024 · similarity 0.85)
- Multi-Modality Affinity Inference for Weakly Supervised 3D Semantic Segmentation (AAAI 2024 · similarity 0.85)
- High Temporal Consistency through Semantic Similarity Propagation in Semi-Supervised Video Semantic Segmentation for Autonomous Flight (CVPR 2025 · similarity 0.85)
- Exploring Scene Affinity for Semi-Supervised LiDAR Semantic Segmentation (CVPR 2025 · similarity 0.84)
- CamPoint: Boosting Point Cloud Segmentation with Virtual Camera (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.