Abstract
Rapid post-event landslide mapping is essential for disaster response but remains difficult to automate due to extreme class imbalance. This study evaluates whether Clay v1.5, a Geo-Foundational Model (GFM), can improve pixel-level landslide segmentation on the Landslide4Sense (L4S) benchmark, which contains 3,799 training chips with 14 Sentinel-2 and terrain bands and approximately 2% positive pixels.
We compare three strategies:
- Clay as the primary encoder with multi-scale residual terrain fusion;
- A U-Net backbone augmented with Clay semantic context at the bottleneck;
- A standard U-Net baseline.
The hybrid U-Net + Clay model with two-stage Low-Rank Adaptation (LoRA) achieved the best test F1 of 64.5 +/- 1.8% over three seeds, surpassing the Clay-only backbone (55.2 +/- 3.6%) and the U-Net baseline (59.9%). While Clay as a standalone encoder underperformed the U-Net due to the absence of multi-scale skip connections, its pretrained representations consistently improved performance when injected as auxiliary context. These findings suggest that GFMs are most effective for landslide detection when they complement spatially detailed convolutional architectures rather than replace them.
Blogger's Review: This study highlights the potential of Geo-Foundational Models in landslide detection, showing significant improvements when combined with U-Net. It not only provides a new perspective for disaster response but also offers important insights for future model designs, emphasizing the importance of multi-scale fusion.