MMLSv2: A Multimodal Dataset for Martian Landslide Detection in Remote Sensing Imagery

TL;DR

MMLSv2 is a comprehensive multimodal dataset designed for landslide detection on Mars, supporting robust model training and evaluation of generalization in remote sensing imagery.

Contribution

The paper introduces MMLSv2, a new multimodal Martian landslide dataset with diverse imagery and an isolated test set to evaluate model robustness and generalization.

Findings

Models trained on MMLSv2 achieve stable performance.

Performance drops on the isolated test set highlight generalization challenges.

The dataset supports research in landslide segmentation and model robustness.

Abstract

We present MMLSv2, a dataset for landslide segmentation on Martian surfaces. MMLSv2 consists of multimodal imagery with seven bands: RGB, digital elevation model, slope, thermal inertia, and grayscale channels. MMLSv2 comprises 664 images distributed across training, validation, and test splits. In addition, an isolated test set of 276 images from a geographically disjoint region from the base dataset is released to evaluate spatial generalization. Experiments conducted with multiple segmentation models show that the dataset supports stable training and achieves competitive performance, while still posing challenges in fragmented, elongated, and small-scale landslide regions. Evaluation on the isolated test set leads to a noticeable performance drop, indicating increased difficulty and highlighting its value for assessing model robustness and generalization beyond standard in-distribution settings. Dataset will be available at: https://github.com/MAIN-Lab/MMLS_v2