The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Understanding Collective Insect Communication with ML, w/ Orit Peleg - #590
Sep 5, 2022
In this discussion, Orit Peleg, an assistant professor at the University of Colorado, Boulder, dives into her fascinating research on insect communication. She shares how swarming behaviors in honeybees mirror principles in distributed computing and emergent systems. Orit also highlights the unique flash patterns of fireflies, showcasing innovative algorithms used to study their synchronization. Delving into conservation, she addresses the impact of light pollution on fireflies and explores how machine learning enhances our understanding of these complex insect societies.
AI Snips
Chapters
Transcript
Episode notes
Firefly Synchronization Discovery
- Synchronous fireflies were first observed in Thailand, sparking debate about the phenomenon's validity.
- Some dismissed it as an illusion caused by observers' blinking or boat movements.
Stereo Vision for Firefly Research
- Orit Peleg's research uses stereo vision with two cameras to capture firefly flashes in 3D space.
- This setup, similar to human depth perception, helps determine the distance between fireflies, a key factor in understanding their communication.
Crowdsourcing Firefly Data
- Orit Peleg's team aims to crowdsource firefly data collection by simplifying camera calibration.
- Volunteers can record fireflies with any two cameras, eliminating complex setup procedures.