Two revolutions have been occurring at the same time in recent years: the citizen science revolution and the artificial intelligence revolution. And revolution is not too strong a term to use for either.

Although one could say that citizen science has been around since the days of Darwin, since he had no formal degrees in his field and pursued his studies of evolution out of personal interest, it has been more or less in the last 5 years or so that citizen science has really come into its own as an academic discipline.

Citizen science or community science, as it is sometimes known, has been using data gathered by lay-persons, under the supervision of scientists, to monitor everything from butterfly and bird populations, to water quality and astronomical anomalies. ¹

And, of course, there is the recent revolution in artificial intelligence or AI as it is commonly known, along with no small amount of hand-wringing about what will happen when computers start driving our cars for us, or worse yet: become smarter than their creators, an event known as the singularity.

iNaturalist and AI

However, a less sinister development would be the marrying of AI and citizen science and in the most recent smartphone app developed by Scott Loarie and his team at iNaturalist and the California Academy of Sciences, using neural networks to compare photographs taken by the user to their database of over 30,000 species of plants and animals. ² Only ‘research grade’ (high quality) photographs are used and the app isn’t perfect because it relies on the quality of the photograph used as well as the number of organism key morphological features visible to the program — so, for example, it is more accurate in identifying the genus of some organisms than it is the species.  In other words, it gives you its ‘best guess’ for fine tuning later.  Nonetheless, as the more data is uploaded to the program, it ‘learns’ and becomes more accurate over time.³ I have used iNaturalist for marine species ID while diving off the coast of California and have yet to be unable to locate a species in the database—even some of the more obscure marine worm species.

Tracking Sharks with Pattern Recognition

The San Diego-based nonprofit Ocean Sanctuaries was founded in 2014 to create and provide support for marine citizen science projects. The Sevengill Shark Identification Project was one of its first citizen science projects, begun in 2010 in response to anecdotal evidence that divers were seeing increasing numbers of Sevengill Sharks off the coast of San Diego.  The IUCN (International Union for Conservation of Nature) has declared this species of shark to be ‘data deficient, making it an ideal candidate for a long-term (5–10 year) population study. During this period, Ocean Sanctuaries partnered with information architect Jason Holmberg, who had helped design the original pattern recognition algorithm that had been used to identify Whale Sharks, (Rhincodon typus).⁴

The way it works is: local divers take lateral head-view photographs of the species during an encounter—when possible without endangering their safety—and, then, submit their photographs to the Wildbook program online at Sevengill Shark Sightings.

Contained within Wildbook are two pattern recognition algorithms, which scan submitted photographs and analyze the black freckling pattern, seen around the eyes and gill area of this species facilitating individual recognition. In this way, photographs submitted over time in a single location can determine which sharks are returning to the location from year-to-year. ⁵

So, it’s no exaggeration to say that AI is revolutionizing citizen science and that citizen science is revolutionizing how the general public’s ability to contribute scientifically reliable data.

About the author

Michael Bear is Citizen Science Project Director for Ocean Sanctuaries, an ocean non-profit dedicated towards supporting and promoting marine citizen science.  He lives and works in San Diego.

You can contact him via LinkedIn: Michael Bear, on Facebook: and Twitter: @rapturedeep

References:

1. Andersen R. 2015 The most mysterious star in our galaxy. The Atlantic; [Online.] www.theatlantic.com/science/archive/2015/10/the-most-interesting-star-in-our-galaxy/410023/

2. Yong E. 2017 Finally: An App That Can Identify the Animal You Saw on Your Hike. The Atlantic; [Online] https://www.theatlantic.com/science/archive/2017/07/an-app-for-identifying-animals-and-plants/535014/

3. Ibid. Above

4. Holmberg, J., Norman, B. and Arzoumanian, Z. (2008), ROBUST, COMPARABLE POPULATION METRICS THROUGH COLLABORATIVE PHOTO-MONITORING OF WHALE SHARKS RHINCODON TYPUS. Ecological Applications, 18: 222–233. doi:10.1890/07-0315.1

5. Peter A. Mieras, Chris Harvey-Clark, Michael Bear, Gina Hodgin, Boone Hodgin, Chapter Five – The Economy of Shark Conservation in the Northeast Pacific: The Role of Ecotourism and Citizen Science, Editor(s): Shawn E. Larson, Dayv Lowry, In Advances in Marine Biology, Academic Press, Volume 78, 2017, Pages 121-153, ISSN 0065-2881, ISBN 9780128123942, https://doi.org/10.1016/bs.amb.2017.08.003. (http://www.sciencedirect.com/science/article/pii/S0065288117300093)