8 Popular Remote Sensing Object Detection Datasets with One-Click Downloads

InteriorGS 3D Indoor Scene Dataset Estimated size: 19.84 GB Download: https://go.hyper.ai/eyG9q Provides high‑quality 3D Gaussian Splatting (3DGS) representations, instance‑level semantic bounding boxes, and occupancy maps indicating navigable areas. Contains 1,000 indoor scenes with floor plans covering residential, convenience store, museum, and other environments (80+ types). Includes >554,000 object instances spanning 755 categories.

Landslide4Sense Landslide Remote Sensing Benchmark Estimated size: 2.84 GB Download: https://go.hyper.ai/mIdeN Multi‑source satellite benchmark released by IARAI in 2022. Covers 2015‑2021, providing 128×128 patches at ~10 m/pixel with 14 spectral bands. Split into 3,799 training, 245 validation, and 800 test samples. Each sample includes a pixel‑level binary mask (landslide = 1). Only training labels are public; validation and test sets are reserved for leaderboard evaluation.

VisDrone Drone Detection Dataset Estimated size: 2.1 GB Download: https://go.hyper.ai/Odzam Large‑scale benchmark from Tianjin University’s AISKYEYE team for object detection, tracking, and segmentation. Contains high‑resolution images and videos captured in Chinese cities and suburbs, covering six categories (people, vehicles, buildings, animals, etc.) under varied lighting, angles, and motion. Annotations are provided in YOLO format.

Crops Disease Agricultural Crop Disease Dataset Estimated size: 1.99 GB Download: https://go.hyper.ai/exltg Contains roughly 1,300 images of common crop diseases across multiple crops (corn, tomato, potato, etc.). Each image is labeled with a specific disease class to support computer‑vision models for automated disease detection and classification.

Safety Vests Detection Dataset Estimated size: 408.58 MB Download: https://go.hyper.ai/ahI1u Includes 3,897 high‑resolution photos of workers wearing or not wearing safety vests. Bounding‑box annotations for “Safety Vest” and “No Safety Vest” are provided under diverse indoor and outdoor conditions, supporting benchmarking of detection architectures such as YOLOv8, Faster‑RCNN, and SSD.

HBFMID Human Fracture Image Dataset Download: https://go.hyper.ai/p7As2 Released by the International University of Bangladesh in 2024. Contains X‑ray and MRI images of various bones (elbow, finger, forearm, humerus, shoulder, femur, tibia, knee, hip, wrist, spine). Original set: 641 images split into 449 training, 128 validation, 64 test. Images are resized to 640×640, contrast‑enhanced, and augmented (flipping, rotation, scaling, cropping, brightness, saturation) to produce a total of 1,539 images.

Urban Issues City Problems Image Dataset Download: https://go.hyper.ai/VqriU Public image classification dataset for recognizing urban infrastructure and environmental problems. Contains ten categories (damaged roads, potholes, illegal parking, etc.). Images are organized per class and annotated in YOLO format, suitable for both classification and object‑detection models.

New Plant Diseases Plant Disease Image Dataset Download: https://go.hyper.ai/C0DhD Published in 2018. Provides ~87,000 RGB images across 38 categories (healthy leaves and various disease types). Split into an 80 % training set, 20 % validation set, and a 33‑image test set. Designed for plant disease recognition, precision agriculture, and related deep‑learning research.