🖥️Try out the Unet trained on VegAnn by yourself with colab :5/5)

colab.research.google.com/drive/1uw7dccO…

🤝This work has been realised with @uq_news @Arvalisofficiel , @UTokyo_News @INRAE_France @Cirad @hiphen_plant and colleagues @kaaviyaVe @guowei_net @MarioSerouart @CAPTE_team ... (4/5)

➡️The images were acquired with different acquisition systems 📷📱 (smartphone, pole-mounted, robot, uav) and under different lighting conditions (3/5)

VegAnn has been constructed to represent diverse crop growing conditions and contains :

➡️3795 images with their corresponding masks
➡️27+ species 🌱
➡️from 200+ location 🌍
(2/5)