Skip to content

D2J3D/Airplanes_detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 

Repository files navigation

Airplanes detection with YOLO

Dataset details: https://universe.roboflow.com/datasets-84fcz/airplane-vgvsf/dataset/1

Dataset description

Dataset contains 4140 images of airplanes. For training dataset we'll take 2898 (70%) images and validate on 828 (20%).

Model

For object detection we will use yolov8s model from Ultralytics

Detection example

val_batch1_pred

Confusion matrix

image

Loss and metrics

image

Planned improvments

As the graphs shows we already have pretty nice results - the model converges and the training might me continued. But there are steel techniques that could be implementec for achieving better results.

  • Use more data augmentation to avoid overfitting
  • Use trained model for more complex task of counting airplanes on the airfield (bird's eye view shots)

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published