During Alaska stream research, teams developed an automated salmon-counting system to track fish in sonar video and reduce reliance on manual counting

Alaska's manual salmon counting techniques are laborious and expensive. To improve efficiency, new computer vision algorithms are under development to automate the vital task of fish population monitoring. Initial AI solutions faced high error rates compared to human counters, but ongoing advancements have significantly enhanced their accuracy. Embracing automation holds the potential for widespread and efficient oversight of crucial fish stock levels.