Estimation of population trends, detection of rare species, and impact assessments\u00a0are important tasks for biologists. Recently, our team had the pleasure of working with Conservation Metrics<\/a>, a services provider for automated wildlife monitoring, on a project to identify red-legged kittiwakes in photos from game cameras. Our work\u00a0included labeling data, model training on the\u00a0Azure Machine Learning Workbench<\/a> platform using Microsoft Cognitive Toolkit (CNTK) and Tensorflow, and deploying a prediction web service.<\/p>\n In this code story, we’ll discuss different aspects of our solution, including:<\/p>\n Below is a\u00a0video<\/a>\u00a0(provided by Abram Fleishman of San Jose State University and Conservation Metrics, Inc<\/a>) capturing the native habitat of red-legged kittiwakes, species that we worked on detecting.\u00a0Biologists use various pieces of equipment, including climbing gear, to install cameras on those cliffs and take daytime and nighttime photos of birds.<\/p>\n\n
<\/a>Data<\/b><\/h2>\n