Inspiration
Wildlife trafficking is an important driver of animal extinctions, and poses some serious threats to the economies, livestock biosecurity and human health worldwide. The ilegal trade of wild species corresponds to the third largest type of traffic in the planet, after drugs and weapons, and has devastating consequences for the species survival and the maintenance of key ecological services, specially when it is associated with habitat reduction and degradation. In addition, trafficked animals increase the risk of new outbreaks of zoonotic diseases because the animals are kept in cages and transported without any concern for sanitary conditions. By protecting wildlife and natural places, we can foster sustainable economies and prevent spillover events.
Only in Brazil, at least 38 million animals are illegally removed from their natural ecosystems, smuggled and sold every year with high profits in the criminal markets. We must counter this horrifying demand for wild animals. We can combat illegal animal trade with the use of technologies to detect and identify species that are trafficked but often go unnoticed by the authorities and most people are unaware of.
Standardized image recognition, structured databases and integrated applications are important tools in the task of detecting animal images to better inform the authorities and raise people awareness about the species targeted by these criminals.
Di Minin, E., Fink, C., Tenkanen, H., & Hiippala, T. (2018). Machine learning for tracking illegal wildlife trade on social media. Nature Ecology & Evolution, 2(3), 406–407. doi:10.1038/s41559-018-0466-x
Daszak, Peter & Olival, Kevin & Li, Hongying. (2020). A strategy to prevent future pandemics similar to the 2019-nCoV outbreak. Biosafety and Health. 2. 10.1016/j.bsheal.2020.01.003.
PREVENTING THE NEXT PANDEMIC - Zoonotic diseases and how tobreak the chain of transmission https://reliefweb.int/sites/reliefweb.int/files/resources/ZP.pdf
What it does
A customizable classifier for image recognition from IBM (Watson Visual Recognition) that identifies species of wild animals, to be used to detect, denounce and monitor trafficked species.
How I built it
Using Watson Studio - IBM.
Challenges I ran into
Our project depends on a large database and a challenge was the deadline.
What I learned
Even with the advancement of technology, fighting animal trafficking is not an easy task, but it is an urgent problem and we all need to mobilize and be aware of the impacts of animal trafficking.
What's next for Projeto Harpia
Improve image recognition and database adding information of life stages and sexual dimorphism, create statistics about trafficked animals by Power BI and cross-validate species identifications with specialists.
Log in or sign up for Devpost to join the conversation.