Research map / Detection and segmentation / Open-set and few-shot detection
LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation
Junjue Wang, Zhuo Zheng, 马爱龙, Xiaoyan Lu, Yanfei Zhong
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-scale training augmentation (ablation) | DeepLabV3+ on LoveDA · LoveDA Test | mIoU (%): 47.62 -> 49.97 | Improved |
| Add multi-scale testing to multi-scale training (ablation) | DeepLabV3+ on LoveDA · LoveDA Test | mIoU (%): 49.97 -> 51.18 | Improved |
| Add multi-scale training augmentation (ablation) | UNet on LoveDA · LoveDA Test | mIoU (%): 48.00 -> 50.21 | Improved |
| Add multi-scale testing to multi-scale training (ablation) | UNet on LoveDA · LoveDA Test | mIoU (%): 50.21 -> 51.13 | Improved |
| Add multi-scale training augmentation (ablation) | Semantic-FPN on LoveDA · LoveDA Test | mIoU (%): 48.15 -> 50.80 | Improved |
| Add multi-scale testing to multi-scale training (ablation) | Semantic-FPN on LoveDA · LoveDA Test | mIoU (%): 50.80 -> 51.82 | Improved |
| Add multi-scale training augmentation (ablation) | HRNet on LoveDA · LoveDA Test | mIoU (%): 49.79 -> 51.51 | Improved |
| Add multi-scale testing to multi-scale training (ablation) | HRNet on LoveDA · LoveDA Test | mIoU (%): 51.51 -> 52.14 | Improved |
| Add background BCE supervision (ablation) | Semantic-FPN on LoveDA · LoveDA Test | mIoU (%): 48.15 ± 0.17 -> 48.85 ± 0.15 | Improved |
| Add background Dice loss to BCE supervision (α=0.5, β=0.5) (ablation) | Semantic-FPN on LoveDA · LoveDA Test | mIoU (%): 48.85 ± 0.15 -> 49.23 ± 0.09 | Improved |
| Add background Dice loss to BCE supervision (α=0.7, β=0.5) (ablation) | Semantic-FPN on LoveDA · LoveDA Test | mIoU (%): 48.74 ± 0.19 -> 50.08 ± 0.15 | Improved |
| Use a larger target class proportion for CBST, t=0.5 rather than t=0.1 (ablation) | CBST, Rural → Urban · Rural → Urban Test | mIoU (%): 36.84 -> 41.32 | Improved |
Closest papers by meaning
- CrossEarth: Geospatial Vision Foundation Model for Domain Generalizable Remote Sensing Semantic Segmentation (TPAMI 2025 · similarity 0.89)
- Towards Open-Vocabulary Remote Sensing Image Semantic Segmentation (AAAI 2025 · similarity 0.88)
- SynRS3D: A Synthetic Dataset for Global 3D Semantic Understanding from Monocular Remote Sensing Imagery (NeurIPS 2024 · similarity 0.87)
- ROS-SAM: High-Quality Interactive Segmentation for Remote Sensing Moving Object (CVPR 2025 · similarity 0.86)
- Learning without Exact Guidance: Updating Large-scale High-resolution Land Cover Maps from Low-resolution Historical Labels (CVPR 2024 · similarity 0.86)
- SemStereo: Semantic-Constrained Stereo Matching Network for Remote Sensing (AAAI 2025 · similarity 0.85)
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.
