The Fishial.AI Project is working on building fishial recognition; open source model that uses machine learning, computer vision and artificial intelligence to correctly identify fish species. Implemented by The Wye Foundation, Fishial.AI seeks to be on the leading forefront of technology to assist in global conservation efforts. Operating as an open source model, fishial recognition is a multifaceted resource offering endless opportunities to help the scientific community, fisheries community and anglers worldwide. With the potential to be incorporated into any platform, scientist, fisheries managers and anglers can create platforms that best suit them to operate fishial recognition powered by Fishial.AI.
Building a successful model to identify fish down to species level will not be an easy task, and will require years of hard work. To conquer such a task the team at Fishial.AI has broken down the project into smaller projects. Since we anticipate the public being such a large part of Fishial.AI we wanted to share our plan.
Milestone 1: Build a front end website that can inform the public of our goal to create fishial recognition. Fishial.AI will also link the public to the citizen science portal.
Milestone 2: Create a citizen science portal. This portal will allow users to create a profile, a team or join a team. Here they will be able to upload their fish images, tag species names and fish attributes and submit the data to the model.
Milestone 3: Build a team on the backend that will verify fish species identifications and fish attributes. Once they are confirmed the images will be sent to the model.
Milestone 4: Train the model.
Milestone 5: Test the model.
Milestones 4 and 5 will need to be repeated several times to ensure that we have built a robust model.
Milestone 6: Offer the model to the scientific community, fisheries community, private sectors and NGO’s as an open source model.