NEW YORK – The Stavros Niarchos Foundation (SNF) Brain Insight Lecture series at Columbia University’s Mortimer B. Zuckerman Mind Brain Behavior Institute continued on February 9 with a fascinating lecture titled ‘A Clear Vision: How Animals and Robots See the Physical World.’ The conversation via Zoom featured Dr. Rudy Behnia and Dr. Shuran Song and was streamed live on the Zuckerman Institute’s YouTube channel.

According to the event’s description, “sensing the world around us might feel effortless, but how does the brain succeed in the complex task of interpreting what we see? Could understanding the neuroscience of vision be the key for creating robots that perceive, interact with, and learn from their surroundings as well as we can – or even better? In this pair of talks by two experts in distinct but related fields, the speakers explored vision, from biology to technology, and discussed how the domains of science and engineering can inspire each other in tackling these fascinating questions.”

Dr. Diana Li, Associate Director of Education & Training Initiatives at the Zuckerman Institute, gave the welcoming remarks, noting that the event would showcase “some fascinating, real-life applications of neuroscience research.” She also mentioned the SNF Teacher Scholar Program which, in conjunction with the lecture series, supports local science teachers bringing the content from the lecture series to their classrooms. Dr. Li thanked SNF and the Foundation members who attended the lecture in person “for their continued partnership and commitment to helping make brain science accessible for all.”

Rudy Behnia, PhD, Assistant Professor of Neuroscience and Principal Investigator at Columbia University’s Zuckerman Institute, opened the event by sharing her research on how the brain processes the dynamic information coming in from our surrounding environment. Using the sophisticated yet compact brains of fruit flies as a model, she brought the audience into the world of animal behavior and discussed the neuroscience behind color perception. Dr. Behnia began her presentation by thanking SNF and then showed a slide of various eyes, highlighting “the diversity of vision that exists in the animal kingdom”

“They [the eyes] were shaped by evolution to best serve the animals in question,” she said and shared some examples of the ways vision in some animals differs from human vision. Though Dr. Behnia began by focusing on the eye, she pointed out that “vision, and perception as a whole, is really a construct of our brain.”

Shuran Song, PhD, Assistant Professor in the Department of Computer Science at Columbia University, then spoke about her work building computer vision algorithms at the forefront of robotics that enable machines to see better. Drawing from examples in her lab, she discussed how robots can sense and perceive more than ever before by actively exploring the physical world and learning from these interactions to carry out complex tasks.

In her presentation, titled ‘Interactive Perception for Embodied Artificial Intelligence,’ Dr. Song noted that embodied AI, robots, “have a physical body that allows them to interact with this environment” and “this really makes all the difference when designing their perception algorithms.”

She then showed video of different robots performing various tasks, and slides to highlight the ways robots perceive their environment in comparison with the ways humans and animals do and then discussed if we can endow robots to share a similar capability “to help them see better.”

Following the two talks, Vassiki Chauhan, PhD, Postdoctoral Fellow at Barnard College, moderated a discussion and Q&A session with the speakers. Audience questions were welcomed, either submitted during registration or live during the event, adding an even greater level on interest to the informative event.

The SNF Brain Insight Lecture ‘A Clear Vision: How Animals and Robots See the Physical World’ is available for viewing online: https://bit.ly/3JWKzGm