HDDS Datasets

#hdds

• Best places to get test files:

1. Best all-in-one source: DroneDetectionThesis dataset
   • Link: https://github.com/DroneDetectionThesis/Drone-detection-dataset
   • Has visible videos, IR videos, and audio clips.
   • Good for your app because it already contains drone/non-drone sensor data.
   • License shown as CC0-1.0, so it is easy to use for testing.
   • Dataset includes visible drone videos and audio labels like Drone, Helicopter, Background.
2. Long drone audio: NASA Small UAS Flyover Acoustics
   • Link: https://data.nasa.gov/dataset/small-uas-flyover-acoustics-data
   • Real small-UAS flyover audio recordings.
   • Good if you want realistic long drone sound.
3. Large drone audio dataset: DADS on Hugging Face
   • Link: https://huggingface.co/datasets/matei219/drone-audio-detection-samples
   • Contains drone and no-drone audio samples.
   • WAV-style audio, 16 kHz mono, clips from short samples to several minutes.
4. Drone video with multiple UAV targets
   • Link: https://engineering.purdue.edu/~bouman/UAV_Dataset/
   • Has 50 video sequences with one or multiple UAVs in view.
   • Useful for multi-target vision testing.

For your app, place files like this:

Scenarios/ scenario-1.mp4 scenario-1.wav scenario-1.json

They do not have to come from the same dataset for simulation testing. You can use a drone video from the GitHub/Purdue dataset, a drone audio file from NASA/DADS, then write your own radar JSON with matching target distance/speed values.