A Fish species detection and identification model made using YOLOv5s, l and x models which can identify fishes of 6 species which are taken from the deepfish.ai dataset and annotated using roboflow. The model aims to detect fishes under poor conditions such as bad lighting, camouflaged instances, murky or cloudy water quality, etc.
The dataset used here is prepared using roboflow from the official website www.roboflow.com
Some of the results for the underwater videos are displayed below:
Fish_Detection_results.mp4
👤 Varun Kalbhore
- Github: @kinkaid002
👤 Abhiyant Gwalani
- Github: @abhiyant-10