AI Paper Atlas

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

TPAMI 2026 · Image generation › Autoregressive and latent generation · approach: Retrieval and matching with generative features · 6 citations (OpenAlex)

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 changedWhereMeasuredEffect
use Vision Transformer architectures instead of DeepLabv3+ or U-NetDL calving front systems · CaFFe test setMDEImproved
add global-scale information via larger inputs and dedicated strategies (deeper U-Net, ASPP, HookNet, attention)DL calving front systems · CaFFe test setMDEImproved
train with zone labels or multi-task learning instead of binary front labels onlyDL calving front systems · CaFFe test setMDEImproved
use foundation model zero-shot instead of prompt-generating modelfoundation model system · CaFFe test setMDEHurt
replace human annotators with best DL system (HookFormer)calving front delineation · CaFFe test setMDE: 38 -> 221Hurt
Sentinel-1 images instead of other sensorsHookFormer · CaFFe test setMDE: 915 vs 105 (TSX)Hurt
winter instead of summer imagesHookFormer · CaFFe test setMDE: 174 -> 271Hurt
Columbia instead of Mapple Glacier imagesHookFormer · CaFFe test setMDE: 110 -> 245Hurt
Mapple instead of Columbia Glacier imageshuman annotators · CaFFe test setMDE: 34 -> 57Hurt
use higher resolution (7 m) instead of 20 m imagesHookFormer · CaFFe test setMDE: 813 -> 105Improved
add conditional random field post-processingbaseline system · CaFFe test setMDEImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy