Deep Learning Pupillometry
A flexible AI pipeline that tracks pupil and eyelid landmarks from video, to understand brain state through the pupil and support brain-wide imaging experiments.
- Program
- Research & FYP
- Monash Automation member
- Jaden Mendis
- Academic advisors
- Daisuke Shimioka, Madhu Permul
Understanding brain state through the pupil
Pupil diameter and eye movement are valuable indicators of arousal, attention and nervous-system activity. Synchronising these measurements with techniques such as widefield calcium imaging, Neuropixels recordings and optogenetic stimulation helps researchers understand how brain-wide activity relates to behaviour and internal state.
A flexible AI pupillometry pipeline
The team is developing a deep-learning pipeline that tracks pupil and eyelid landmarks from video to measure pupil diameter, position and movement over time. Designed for high-fidelity, low-latency analysis, it can support different neuroscience experiments and may ultimately enable real-time, closed-loop systems that respond to detected changes in brain state.
