RarePlanes

Stream data with DDA:

from dagshub.streaming import DagsHubFilesystem

fs = DagsHubFilesystem(".", repo_url="https://dagshub.com/DagsHub-Datasets/rareplanes-dataset")

fs.listdir("s3://rareplanes-public")

Description:

RarePlanes is a unique open-source machine learning dataset from CosmiQ Works and AI.Reverie that incorporates both real and synthetically generated satellite imagery. The RarePlanes dataset specifically focuses on the value of AI.Reverie synthetic data to aid computer vision algorithms in their ability to automatically detect aircraft and their attributes in satellite imagery. Although other synthetic/real combination datasets exist, RarePlanes is the largest openly-available very high resolution dataset built to test the value of synthetic data from an overhead perspective. The real portion of the dataset consists of 253 Maxar WorldView-3 satellite scenes spanning 112 locations and 2,142 km^2 with 14,700 hand-annotated aircraft. The accompanying synthetic dataset is generated via AI.Reverie’s novel simulation platform and features 50,000 synthetic satellite images with ~630,000 aircraft annotations.

Contact:

RarePlanes is a unique open-source machine learning dataset from CosmiQ Works and AI.Reverie that incorporates both real and synthetically generated satellite imagery. The RarePlanes dataset specifically focuses on the value of AI.Reverie synthetic data to aid computer vision algorithms in their ability to automatically detect aircraft and their attributes in satellite imagery. Although other synthetic/real combination datasets exist, RarePlanes is the largest openly-available very high resolution dataset built to test the value of synthetic data from an overhead perspective. The real portion of the dataset consists of 253 Maxar WorldView-3 satellite scenes spanning 112 locations and 2,142 km^2 with 14,700 hand-annotated aircraft. The accompanying synthetic dataset is generated via AI.Reverie’s novel simulation platform and features 50,000 synthetic satellite images with ~630,000 aircraft annotations.

Update Frequency:

None Planned

Managed By:

In-Q-Tel - CosmiQ Works

Resources:

1. resource:
   • Description: Real and synthetic satellite imagery, annotations, and metadata
   • ARN: arn:aws:s3:::rareplanes-public
   • Region: us-west-2
   • Type: S3 Bucket

Tags:

computer vision, deep learning, earth observation, geospatial, machine learning, satellite imagery, aws-pds, labeled

Tutorials:

1. tutorial:

   • Title: Announcing YOLTv4: Improved Satellite Imagery Object Detection
   • URL: https://towardsdatascience.com/announcing-yoltv4-improved-satellite-imagery-object-detection-f5091e913fad
   • AuthorName: Adam Van Etten
   • AuthorURL: https://medium.com/@avanetten

2. tutorial:

   • Title: Getting Started with YOLTv4 for Object Detection in Imagery: Getting Training Data
   • URL: https://dev.to/spara_50/getting-started-with-yoltv4-for-object-detection-in-imagery-getting-training-data-2055
   • AuthorName: Sophia Parafina
   • AuthorURL: https://authory.com/sophiaparafina

3. tutorial:

   • Title: Notebook for training and testing YOLYv4
   • URL: https://github.com/avanetten/yoltv4/blob/master/notebooks/train_test_pipeline.ipynb
   • AuthorName: Adam Van Etten

Tools & Applications:

1. tools & applications:
   • Title: RarePlanes Codebase
   • URL: https://github.com/aireveries/RarePlanes
   • AuthorName: Thomas Hossler and Jacob Shermeyer

Publication:

1. publication:
   • Title: RarePlanes: Synthetic Data Takes Flight
   • URL: https://arxiv.org/abs/2006.02963
   • AuthorName: Jacob Shermeyer, Thomas Hossler, Adam Van Etten, Daniel Hogan, Ryan Lewis, Daeil Kim