Mirali33/mb-landmark_cls

mb-landmark_cls

A Mars image classification dataset for planetary science research.

Dataset Metadata

• —License: CC-BY-4.0 (Creative Commons Attribution 4.0 International)
• —Version: 1.0
• —Date Published: 2025-05-14
• —Cite As: TBD

Classes

This dataset contains the following classes:

• —0: oth
• —1: cra
• —2: ddu
• —3: sst
• —4: bdu
• —5: ime
• —6: sch
• —7: spi

Statistics

• —train: 6997 images
• —test: 1793 images
• —val: 2025 images
• —few_shot_train_2_shot: 16 images
• —few_shot_train_1_shot: 8 images
• —few_shot_train_10_shot: 80 images
• —few_shot_train_5_shot: 40 images
• —few_shot_train_15_shot: 120 images
• —few_shot_train_20_shot: 160 images
• —partition_train_0.01x_partition: 69 images
• —partition_train_0.02x_partition: 139 images
• —partition_train_0.50x_partition: 2646 images
• —partition_train_0.20x_partition: 1286 images
• —partition_train_0.05x_partition: 349 images
• —partition_train_0.10x_partition: 692 images
• —partition_train_0.25x_partition: 1570 images

Few-shot Splits

This dataset includes the following few-shot training splits:

• —few_shot_train_2_shot: 16 images
• —few_shot_train_1_shot: 8 images
• —few_shot_train_10_shot: 80 images
• —few_shot_train_5_shot: 40 images
• —few_shot_train_15_shot: 120 images
• —few_shot_train_20_shot: 160 images

Few-shot configurations:

• —2_shot.csv
• —1_shot.csv
• —10_shot.csv
• —5_shot.csv
• —15_shot.csv
• —20_shot.csv

Partition Splits

This dataset includes the following training data partitions:

• —partition_train_0.01x_partition: 69 images
• —partition_train_0.02x_partition: 139 images
• —partition_train_0.50x_partition: 2646 images
• —partition_train_0.20x_partition: 1286 images
• —partition_train_0.05x_partition: 349 images
• —partition_train_0.10x_partition: 692 images
• —partition_train_0.25x_partition: 1570 images

Usage

from datasets import load_dataset

dataset = load_dataset("Mirali33/mb-landmark_cls")

Format

Each example in the dataset has the following format:

{
  'image': Image(...),  # PIL image
  'label': int,         # Class label
}