Research & Publications

I am currently focusing on developing light and robust algorithms for low power computational eyeglasses. My ultimate goal is to be able to extract useful information regarding the cognitive state of the user by means of computational eyeglasses.


We present W!NCE, an EOG-based system that can continuously and reliably monitor facial expressions in the natural environment by using a commercially available eyeglass ( J!NS MEME). W!NCE detects facial gestures using a two-stage processing pipeline involving motion artifact removal and facial action detection, which has been extensively evaluated by 17 users separately under stationary and ambulatory settings. We also validate our system’s applicability to continuous pain monitoring through a pilot study and several performance benchmarks.

W!NCE: Unobtrusive Sensing of Upper Facial Action Units with EOG-based Eyewear
Soha Rostaminia, Alexander Lamson, Subhransu Maji, Tauhidur Rahman, Deepak Ganesan
Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp '19)


iShadow is a computational eyeglass that provides a suite of context-sensing tools in a low-power, robust manner, and in an unobtrusive form factor.
We have developed a computationally light algorithm, called iLid, for the purpose of fatigue detection on iShadow platform. iLid can extract eye features such as blink rate, blink duration, and percentage of eye closure (PERCLOS) and can run continuously and up to 2 days using a 570mAh battery, such as the type used on Google Glass.

iLid: Low-power Sensing of Fatigue and Drowsiness Measures on a Computational Eyeglass
Soha Rostaminia, Addison Mayberry, Benjamin Marlin, Deepak Ganesan, Jeremy Gummeson
Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp '17)


In this project, the idea is to instrument the typical cloth sleep mask with textile-based sensors in order to measure eye movement patterns during sleep and improving understanding of sleep stages and circadan rhythm.

iMask prototype


WearID is a wearable RFID reader to sense RFID tags on objects. In this project, we present the design and implementation of a full hardware prototype of such a wearable RF reader that tackles several challenges including reader power consumption and body blockage.

WearID: A Wrist-worn Backscatter Reader to Measure Interactions with the Physical World
Ali Kiaghadi, Pan Hu, Jeremy Gummeson, Soha Rostaminia, Deepak Ganesan
(Under Review)

WearID prototype

Parkinson's Disease Eye Symptoms

In this research, we want to capture and analyze the eye movement/closure patterns of the Parkinson’s patients during the time interval of their morning and afternoon dopaminergic medication intake. The goal is to explore the Parkinson's patients’ eye movement patterns throughout their off and on medication stages. We hope that this investigation leads to finding out the correlations between the patients’ dopamine level and their eye movement behavior.

Parkinson's patient