The North Atlantic Right Whale Catalog is the brainchild of marine biologist Christin Khan, a biologist at the National Oceanic and Atmospheric Administration’s Northeast Fisheries Science Center, who wanted to create an analogue to Facebook for the endangered North Atlantic Right Whale, whose numbers in 2016 hovered at approximately 500. Khan takes up to 60 trips each year to America’s Northeast coast, taking aerial photos from a helicopter of these endangered whales, but wanted a more efficient way to identify the whales.
Khan launched a challenge on Kaggle, a crowdsourcing platform that hosts competitions for researchers and scientists to create innovative solutions to important problems The challenge: successfully automate the identification of North Atlantic Right Whales. An algorithm was eventually created by deepsense.io, a Portland-based company that focuses on innovative data science solutions. The algorithms could correctly identify a whale based on its unique assortment of whitish head markings (called callosities) 87% of the time. The algorithm works by standardizing the aerial photos: first, rotating the photos, so that the blowhole and tip of the whale are in a particular position, and then cropping the photos to a standard size. Next, a trained neural network searches for patterns in callosities, which can then be matched to known patterns of identified (and sometimes named) whales.
Says Kahn in a recent article: “Recent advances in deep learning coupled with this fully automated workflow have yielded impressive results and have the potential to revolutionize traditional methods for the collection of data on the abundance and distribution of wild populations.” Khan and her colleagues hope this approach will help to bridge the gap between the data science and conservation science communities.
Bogucki, Robert, Marek Cygan, Christin Brangwynne Khan, Maciej Klimek, Jan Kanty Milczek, and Marcin Mucha. “Applying deep learning to right whale photo identification.” Conservation Biology 33, no. 3 (2019): 676-684.
See also:
Preston, Elizabeth. (2016, Jan. 14). “Making Facebook for Whales.” The Atlantic. https://www.theatlantic.com/science/archive/2016/01/north-atlantic-right-whales-facial-recognition/424113/
Categories
Artificial Intelligence, Biodiversity, Citizen Science, Ecological Monitoring, Visual Technologies
Air Pollution Robot
The dangers of air pollution to human health are well documented, though the traditional methodology of collecting and reporting on sample lags behind the need to keep abreast and regulate air pollution in a meaningful amount of time. The use of drones and robots have been identified by researchers as resources that can be tweaked […]
Artificial Life, Biodiversity, Climate Change, Ecological Monitoring, Industry/Natural Commodities, Lifestyle, Monitoring
Telematic Rivers
Erica Kermani’s artwork seeks to answer a central question: if rivers were seeing an equal, living entity, would humans take issues like climate change threatening them more seriously? In his year-long art exhibition in 2017, Kermani, in collaboration with Diana Salcedo & Jeana Chesnik, created a new forum of interaction between humans and rivers to […]
Climate Change, Ecological Monitoring, Lifestyle, Monitoring, Pollution, Visual Technologies
Co-occupied Boundaries
Art is easily found in nature but rarely is what considered art today inherently natural. The concept of co-occupied mediums that serve to be both functional for nature and aesthetically pleasing to people is being actively explored by Asya Ilgun and Phil Ayres, from the CITAstudio at The Royal Danish Academy of Fine Arts. In […]
Climate Change, Ecological Monitoring, Lifestyle, Monitoring, Pollution, Visual Technologies