Skip to content

Dataset · RiceD2KLab

SWiM spacecraft segmentation dataset

RiceD2KLab/SWiM-SpacecraftWithMasks

About 64,000 spacecraft images with instance segmentation masks based on a NASA pipeline.

What's inside

The dataset contains over 63,917 spacecraft images with instance segmentation annotations. The images were created by overlaying real spacecraft models onto real and synthetic backgrounds using the NASA TTALOS pipeline, with added noise and camera distortions to simulate real footage.

What you can do with it

  • Training segmentation and detection models for orbital space objects
  • Testing computer vision robustness against noise and optical distortions
Tasks
Image segmentation
Downloads this month
26K
Likes
3

How to load it

from datasets import load_dataset

ds = load_dataset("RiceD2KLab/SWiM-SpacecraftWithMasks")