We developed a new machine learning-based feature for Salmonsoft to automate the process of fish counting at fish viewing windows at fish ladders and weirs. This software has the potential to both greatly reduce personnel time required to count fish and reduce errors. The initial version of the technology also estimates fish length and summarizes length data.

The solution was initially unveiled at the Fish Passage Conference in June of 2016 at the University of Massachusetts in Amherst.’s fish counting solution is an enhancement to Salmonsoft’s FishTick software, which is designed to remove video where no fish pass and also gives users an easy-to-use interface to review the remaining fish passage video. In addition, FishTick will automatically record the date and time of every fish passage event and automatically summarize fish count data. FishTick has been used at over 100 sites across North America and Europe.


Other cases

We developed an anomaly detection and machine learning based system for early detection of anomalies in the performance of Foursquare’s data infrastructure and systems.

Read more

Our team was involved in the design and implementation of MedNition’s machine learning framework, worked with diverse anonymized patient data and used a variety of medical information ontologies.

Read more