Mars-Bench: mb-conequest_det

This is the object detection version of the ConeQuest dataset, formulated to localize cones on the Martian surface using CTX imagery. It was developed to support global mapping and morphologic analysis of small-scale volcanic landforms. The dataset spans six geographically diverse regions on Mars, capturing substantial variation in cone shape, size, and appearance, making it a challenging benchmark for model generalization. Each sample consists of an image and its bounding boxes, with all annotations created by expert geologists to ensure scientific accuracy. As the original ConeQuest contains negative samples, we have removed it from this version of the dataset as many detection models do not support training with image samples that do not have any objects. The dataset comes with pre-defined train, validation, and test splits. To explore all Mars-Bench resources, visit our project website: https://mars-bench.github.io/