System of deep neural networks for recognition of plants and fungi from images, developed by Milan Sulc and Jiri Matas from the Dept. of Cybernetics FEE CTU in cooperation with Lukas Picek from the Dept. of Cybernetics FAS UWB, scored first in three international challenges:
- 2018 FGCVx Flower Classification Challenge
- 2018 FGCVx Fungi Classification Challenge
- LifeCLEF 2018 Expert
In a comparison with human experts in plant recognition (as part of the LifeCLEF 2018 workshop) the system scored better than 5 of 9 experts.
The results will be presented at corresponding workshops of the prestigious conferences – CVPR 2018 and CLEF 2018.