Inspiration
We can help mitigate endangerment of species and promote preservation of species.
What it does
A software tool that will scan an image input, containing a fish, and identify the species of that fish. Our approach is to make identification of fish species easier and more efficient. Our model is one that can be incorporated into a fisherman feedback app developed by INCOIS.
How we built it
The software model is developed using deep learning methods and neural networks, hence it can learn how to identify better with every training iteration.
Challenges faced
The main problem in every deep neural networks is overfitting and the network trys to memorize the stuff but we tried to overcome that problem by making our model to learn not to memorize the things
Accomplishments that we're proud of
We have acquired 93.15 % accuracy in indentifying the species and we have also bulid a web application on it
What we learned
We learnt how to be more productive by managing college and doing this project and along with we also learned new skills like CNN, Computer Vision and Frontend development
What's next for Automated identification of fish species
We can deploy this model into Fishermen feedback app and also we can add more datasets and make a global application
Log in or sign up for Devpost to join the conversation.