Building an AI Model one Species At a Time

We are building a deep neural network to teach machines how to identify over 33,000 fish species to help the scientific, educational, governmental, and private communities in creating more accurate data to save our fisheries.

bg about
bg item square white
grid_2x3grid_2x3fish 200×200grid_3x1grid_3x3grid_1x3grid_1x4grid_2x1fish 200×200




Contribute to Fishial Recognition™️ by sharing your fish images with Fishial.AI!

In order to build Fishial Recognition™️ Fishial.AI needs a large collection of fish images. One or 10,000 images, every bit helps.

Images from fishing or underwater photography may be submitted. If there is a fish in it, we need it.

Images submitted will be used for the purpose of building the ML model only, and other individuals will not have access or rights to your images unless you have given them access via the image creative commons setting.

  1. Machine Learning
  2. Photo Collection
bg item square bluebg oval item

Machine Learning

Creating Deep Neural Networks

The Fishial Recognition™️ model will only be as good as the data that is used to train and teach the model. Therefore, the team at Fishial.AI has decided that the more we know about each fish image submitted the more the model will learn. Fish images used to teach and train the model will have a polygon around the fish, this will minimize background noise. The fish images will also be tagged with identifying attributes so that we can teach the model the difference between species. This will facilitate deep neural networks within the model that will help produce a highly accurate fishial identification tool.


Fishial.Ai wants to thank our partnerships for continued support of this project.

Please visit their websites to learn more about them.

bg item square bluebg oval item


Follow along with the latest news from Fishial.AI on Fishial Recognition™️




Questions can be emailed to

View Story


Fishial.AI Project Overview

Fishial.AI is an open source model that uses machine learning, computer vision and artificial intelligence to correctly identify fish species.

View Story