top of page
Seagulls

Welcome to Wingbeat Watch

Developing a system for automatic detection of bird species from flight pattern

The aim of this project is to develop a system that can identify the species of a bird from its flight pattern. To do this, a dataset of videos of birds in flight is required, labelled by species. Please find out more about the project and contribute to this dataset below.

​

About

The final system will take an input video, split it into frames, and identify whether there is a bird in each frame and the action of the bird in each frame. This will generate a flight pattern, which will be used to identify the species. 

I am a master's student studying engineering at the University of Cambridge, and I am working on this system for my final year project, though I hope to eventually make it available for hobbyist use.

If you have further questions, feel free to send me an email!

wingbeatwatch@gmail.com

Video upload

By uploading your videos to this website, please bear in mind that you are confirming that the videos are your own property and you are consenting to the use of your videos in the dataset. The dataset will not be shared anywhere, and will only be used to train a neural network for machine learning purposes. 

 

See example videos below (video quality isn't very important as long as the bird is visible):

Please note only the first three fields, and the consent checkbox, are required. 

Upload Video
E.g. bird.mp4

Thanks for uploading!

Wingbeat Watch

Thanks for subscribing!

Thank you for visiting!

©2022 by Wingbeat Watch. Proudly created with Wix.com

bottom of page