Machine Learning challenge to classify and segment kerogen macerals in thin rock sections, using labeled and unlabeled images. Focused on identifying liptinite, vitrinite, and inertinite types, the competition emphasizes model accuracy measured by the Jaccard Score to enhance sedimentary rock analysis and hydrocarbon exploration.