MIDS Capstone Project Spring 2019


At a time in human history when the rate of animal extinctions is increasing exponentially, we are still living with season long lags for identifying species counts and seasonal ranges. Stationary camera traps provide a critical tool for ecologists working in this sector, however as more and more camera traps come online, manually labeling camera images is no longer an option. Luckily, we are at an inflection point in machine vision that can enable the automatic classification of these pictures.

At CritterCounter our talented team of data scientists and AI experts are fixing the animal identification bottleneck by using existing labeled wildlife images to build a machine vision model which identifies the species in every shot.  Working with wildlife researchers in Washington State, we are taking out the major pain point in their identification tasks,letting the wildlife biologists spend more time actually looking at animals, and giving the researchers the ability to make statistical inferences at scale.

Heat map generated by CritterCounter
Heat map generated by CritterCounter

Last updated:

October 1, 2019