A Benchmark Study of Segmentation Models and Adaptation Strategies for Landslide Detection from Satellite Imagery

Md Kowsher, Weiwei Zhan, Chen Chen · 2026

Read A Benchmark Study of Segmentation Models and Adaptation Strategies for Landslide Detection from Satellite Imagery by Md Kowsher, Weiwei Zhan, Chen Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-16663.txt.

Description

Kowsher and colleagues benchmark mask models and plans for landslide finding from imagery. Landslide finding matters for help and risk review, while the designs and tuning plans remain unclear. Using the GDCLD dataset, they test conv networks, transformer models and pre-trained base models under rules. They compare full fine-tuning with light methods like LoRA and AdaLoRA. Transformer models achieve masks, while light tuning cuts tuned weights by up to 95 percent with close scores. The study weighs in-house and held test scores to probe transfer under shift.

Catalog record

CREATORMd Kowsher, Weiwei Zhan, Chen Chen
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE19959
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.16663

Browse the full library · About the Tower of Babel