Detecting Social Anxiety Disorder (SAD) scalably is challenging due to the absence of deterministic objective markers. The detection is exacerbated due to the comorbidities. Numerous studies have studied SAD in developed countries under varied contexts. However, only a few related studies are from India, a developing country.
To build scalable diagnostic solutions, we started with the discovery of biomarkers of SAD in Indian settings. We conducted a controlled study with university student participants. The participants performed different anxiety-provoking activities, and during the study period, we collected their ECG, PPG, and GSR with several sensors. Several significant findings from this project are still under review.
Related to this project, following sub-works have been completed
Buildings consume 40% of available energy, with up to 20% wasted due to various reasons such as appliance faults, misconfiguration, and abnormal user behavior. Providing energy feedback can save up to 12%. Our contributions in this area includes
Occupancy detection and counting are crucial to automate buildings’ lighting, heating, and cooling systems. However, most existing occupancy detection and counting techniques are underperforming or privacy-averse. In this work, we propose an adaptive methods for processing thermal images and evaluate several classification algorithms for occupancy counting.
In the transportation domain, our goal is to leverage cheap smartphone sensors in Indian settings for (i) Automated traffic sign detection and identification - This is a challenging problem as Indian roads have a mix of traffic and advertisement boards in English and vernacular languages. (ii) Rash driving detection. Following are the accepted papers from this work.