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

TL;DR

This study systematically benchmarks various segmentation models and fine-tuning strategies for landslide detection in satellite imagery, highlighting the effectiveness of transformer models and parameter-efficient methods.

Contribution

It provides a comprehensive comparison of CNNs, transformers, and foundation models, along with fine-tuning techniques, for landslide detection using a new benchmark dataset.

Findings

Transformer models outperform CNNs in segmentation accuracy.

Parameter-efficient fine-tuning reduces trainable parameters by up to 95%.

Fine-tuning methods maintain accuracy under distribution shifts.

Abstract

Landslide detection from high resolution satellite imagery is a critical task for disaster response and risk assessment, yet the relative effectiveness of modern segmentation architectures and finetuning strategies for this problem remains insufficiently understood. In this work, we present a systematic benchmarking study of convolutional neural networks, transformer based segmentation models, and large pre-trained foundation models for landslide detection. Using the Globally Distributed Coseismic Landslide Dataset (GDCLD) dataset, we evaluate representative CNN- and transformer-based segmentation models alongside large pretrained foundation models under consistent training and evaluation protocols. In addition, we compare full fine-tuning with parameter-efficient fine-tuning methods, including LoRA and AdaLoRA, to assess their performance efficiency tradeoffs. Experimental results show that transformer-based models achieve strong segmentation performance, while parameter efficient finetuning reduces trainable parameters by up to 95% with comparable accuracy to full finetuning. We further analyze generalization under distribution shift by comparing validation and held-out test performance.