Wild Me

Wild Me is a 501(c)(3) non-profit organization focused on wildlife conservation using machine learning and software through the organization's Wildbook platform, a web-based collaboration to generate new population metrics originally applied to whale sharks. The early work done through the Wildbook platform was done through whaleshark.org, which generated interest in applying the Wildbook platform to other species. Since then, the platform has been applied to the study of zebras, humpback whales, sevengill sharks, ragged tooth sharks, and giant manta rays.

With the increased interest in the Wildbook platform, Wild Me made the software open source and built collaborations with researchers to grow the Wildbook platform's data science and computer vision capabilities geared towards new species. The organization maintains the original focus on sharks and rays.

The data collected by the platform works to identify a specific animal through analysis of the animal's unique markings, including stripes and spots; defining physical features, such as scars; and the location, time, and date an image is recorded. As more citizen scientists and researchers add images to the catalog, the data grows and allows wildlife experts to track the health, migration, and important facts about a species. As well, scientists and researchers can use the information provided by the Wildbook platform for climate, geographic, behavioral, and environmental data to better understand the ecological and conservation issues facing animals.

The Wildbook platform is a Java app that uses a templating language called JSP for the front end. As well, the platform includes the Wildbook IA, which is short for Wildbook Image Analysis, and is a machine learning and computer vision stack with a Flask interace. These repositories are available through Wild Me. Wild Me also uses Microsoft Azure and related services, including Azure's Translator Text API, to translate multilingual videos and translate Wildbook for non-English speakers; the Computer Vision API for optical character recognition and for date and location prediction; Azure virtual machines on which Wildbook runs; Data Science Virtual Machines to create natural language processing predictors; Application Insights to monitor Wild Me servers and services for outages; a Container Registry to store Wildbook containers; Azure Backup to back up virtual machines and Wildbook's data; and DevOps to monitor operations during development.

The different iterations, or projects, the Wildbook platform has been used for include:

• Flukebook
• whaleshark.org
• MantaMatcher
• GiraffeSpotter
• Internet of Turtles
• Wildbook for Zebras
• Wildbook for Lynx
• Giant Sea Bass
• Spot a Shark
• African Carnivores
• Spot a Shark USA
• Whiskerbook
• Skunk Wildbook
• Seadragonresearch
• Saimaa Seal
• Leopard Sharks

Financial and material supporters of Wild Me and the Wildbook platform have included:

• The Gordon and Betty Moore Foundation
• H2O.ai
• Microsoft
• NVIDIA
• Marine Megafauna Foundation
• National Oceanic and Atmospheric Administration (NOAA)
• SeadragonSearch
• Lokalise
• Transloadit
• Flatfile
• Algolia
• Pineapple Fund
• National Science Foundation
• PBNF
• Adobe
• Siren Fleet
• British Ecological Society
• Waitt Foundation

Wild Me works with users to find solutions and works with the conservation community to help efforts in conservation and tracking. To help with this, Wild Me offers free and paid services, which includes consultation on technological advancements for mark recapture and citizen scientists projects; Azure cloud hosting of Wildbooks; data upload assistance; annotation of photographic data; configuration and support of Wildbooks; custom computer vision and machine learning model development; and Wildbook feature development.