AI for social good: Saving the planet through data science

Global environmental challenges have pushed our planet to the brink of disaster. Rapid advances in deep learning are placing immense power in the hands of consumers and enterprises. Ganes Kesari explains how this power can be marshaled to support environmental groups and researchers who need immediate assistance to address the rapid depletion of our rich biodiversity.

Ganes focuses on the applications of deep learning to solve such environmental problems. These include detection of fish in underwater videos or camera traps in forests, classification of endangered species from crowdsourced pictures or drone footage in Africa, and counting of penguins from periodic pictures over the years. Along the way, Ganes shares live case studies from the work done for NGOs around the world, as well as implementation challenges and techniques used to overcome them, and leads a demo of the public API libraries built by Gramener in partnership with the Microsoft AI for Earth team to show how deep learning is saving our world, paradoxically, from our own hands.