Clay-CNN Hybrids: Leveraging Geo-Foundational Models as Auxiliary Context for Landslide Detection
arXiv:2606.14081v1 Announce Type: cross 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
The increasing frequency of extreme weather events and the advancements in AI, particularly foundational models, are converging to create opportunities for more automated and effective disaster response.
This development indicates a tangible application of advanced AI for real-world disaster management, potentially saving lives and resources through rapid damage assessment.
The effective integration of geo-foundational models like Clay with CNNs could significantly improve the accuracy and speed of identifying critical geological events.
- · Disaster response agencies
- · Geospatial AI developers
- · Satellite data providers
- · Communities in disaster-prone areas
- · Traditional manual damage assessment methods
Faster and more accurate mapping of landslide events for immediate humanitarian and infrastructure response.
Reduced economic impact from landslides due to improved preparedness and faster recovery efforts.
The establishment of geo-foundational models as critical infrastructure for global environmental monitoring and risk assessment.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI