Dr. Pritish Chakravarty

I am a multidisciplinary researcher interested in the measurement and analysis of animal behavior. I trained as an engineer with a specialization in signal processing, algorithm development, and wearable sensors, and transitioned into animal ecology for my PhD.

I use sensors (e.g. accelerometers, magnetometers, gyroscopes) and/or cameras to record animal behavioral data, typically at high temporal resolution, and analyze these by developing new methods that are based on concepts and tools from signal processing, machine learning, image processing, and biomechanics. These analyses are driven by ecological research questions, where I collaborate closely with behavioral ecologists.

Some past and ongoing projects include: fine-scale foraging strategies in meerkats, movement patterns of leopard tortoises, and perception and motion of fish in visually noisy environments.

My current work focuses on studying sleep in social animal species living in their natural habitat using accelerometers*.

_{*The accelerometer is a versatile (and tiny) motion sensor that is sensitive to both motion and gravity. Roughly speaking, a uniaxial accelerometer is like a block attached to a spring; the block is free to move back and forth along a straight line depending on the mechanical forces acting upon it. The block's displacement is translated to a measurable voltage via specially designed electric circuits. You might have a triaxial accelerometer (with three axes pointing in the up-down, left-right, and forward-backward directions) in your hand right at this very moment, for smartphones today use these sensors to detect when the phone has been picked up or rotated, or to count the number of steps you walked today.}