Research map / Image generation / Autoregressive and latent generation
Comparison Study: Glacier Calving Front Delineation in Synthetic Aperture Radar Images With Deep Learning
Nora Gourmelon, Konrad Heidler, Erik Loebel, Daniel Cheng, Julian Klink, Anda Dong, Fei Wu, Noah Maul, Moritz Koch, Marcel Dreier, D. R. Pyles, Thorsten Seehaus and 3 more
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 |
|---|---|---|---|
| use Vision Transformer architectures instead of DeepLabv3+ or U-Net | DL calving front systems · CaFFe test set | MDE | Improved |
| add global-scale information via larger inputs and dedicated strategies (deeper U-Net, ASPP, HookNet, attention) | DL calving front systems · CaFFe test set | MDE | Improved |
| train with zone labels or multi-task learning instead of binary front labels only | DL calving front systems · CaFFe test set | MDE | Improved |
| use foundation model zero-shot instead of prompt-generating model | foundation model system · CaFFe test set | MDE | Hurt |
| replace human annotators with best DL system (HookFormer) | calving front delineation · CaFFe test set | MDE: 38 -> 221 | Hurt |
| Sentinel-1 images instead of other sensors | HookFormer · CaFFe test set | MDE: 915 vs 105 (TSX) | Hurt |
| winter instead of summer images | HookFormer · CaFFe test set | MDE: 174 -> 271 | Hurt |
| Columbia instead of Mapple Glacier images | HookFormer · CaFFe test set | MDE: 110 -> 245 | Hurt |
| Mapple instead of Columbia Glacier images | human annotators · CaFFe test set | MDE: 34 -> 57 | Hurt |
| use higher resolution (7 m) instead of 20 m images | HookFormer · CaFFe test set | MDE: 813 -> 105 | Improved |
| add conditional random field post-processing | baseline system · CaFFe test set | MDE | Improved |
Closest papers by meaning
- Statistically Principled Deep Learning for SAR Image Segmentation (AAAI 2024 · similarity 0.81)
- Kuro Siwo: 33 billion $m^2$ under the water. A global multi-temporal satellite dataset for rapid flood mapping (NeurIPS 2024 · similarity 0.79)
- Galileo: Learning Global & Local Features of Many Remote Sensing Modalities (ICML 2025 · similarity 0.79)
- IceDiff: High Resolution and High-Quality Arctic Sea Ice Forecasting with Generative Diffusion Prior (CVPR 2025 · similarity 0.79)
- Benchmarking Object Detectors under Real-World Distribution Shifts in Satellite Imagery (CVPR 2025 · similarity 0.79)
- A General Adaptive Dual-level Weighting Mechanism for Remote Sensing Pansharpening (CVPR 2025 · similarity 0.79)
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.
