EyeTrAES: Fine-grained, Low-Latency Eye Tracking via Adaptive Event Slicing

Eye-tracking technology has gained significant attention in recent years due to its wide range of applications in human-computer interaction, virtual and augmented reality, and wearable health. Traditional RGB camera-based eye-tracking systems often struggle with poor temporal resolution and computational constraints, limiting their effectiveness in capturing rapid eye movements. To address these limitations, we propose EyeTrAES, a novel approach using neuromorphic event cameras for high-fidelity tracking of natural pupillary movement that shows significant kinematic variance. One of EyeTrAES's highlights is the use of a novel adaptive windowing/slicing algorithm that ensures just the right amount of descriptive asynchronous event data accumulation within an event frame, across a wide range of eye movement patterns. EyeTrAES then applies lightweight image processing functions over accumulated event frames from just a single eye to perform pupil segmentation and tracking. We show that these methods boost pupil tracking fidelity by 6+%, achieving IoU~=92%, while incurring at least 3x lower latency than competing pure event-based eye tracking alternatives [38]. We additionally demonstrate that the microscopic pupillary motion captured by EyeTrAES exhibits distinctive variations across individuals and can thus serve as a biometric fingerprint. For robust user authentication, we train a lightweight per-user Random Forest classifier using a novel feature vector of short-term pupillary kinematics, comprising a sliding window of pupil (location, velocity, acceleration) triples. Experimental studies with two different datasets demonstrate that the EyeTrAES-based authentication technique can simultaneously achieve high authentication accuracy (~=0.82) and low processing latency (~=12ms), and significantly outperform multiple state-of-the-art competitive baselines.

AirSPEC: An IoT-empowered Air Quality Monitoring System integrated with a Machine Learning Framework to Detect and Predict defined Air Quality parameters

The air that surrounds us is the cardinal source of respiration of all life-forms. Therefore, it is undoubtedly vital to highlight that balanced air quality is utmost important to the respiratory health of all living beings, environmental homeostasis, and even economical equilibrium. Nevertheless, a gradual deterioration of air quality has been observed in the last few decades, due to the continuous increment of polluted emissions from automobiles and industries into the atmosphere. Even though many people have scarcely acknowledged the depth of the problem, the persistent efforts of determined parties, including the World Health Organization, have consistently pushed the boundaries for a qualitatively better global air homeostasis, by facilitating technology-driven initiatives to timely detect and predict air quality in regional and global scales. However, the existing frameworks for air quality monitoring lack the capability of real-time responsiveness and flexible semantic distribution. In this paper, a novel Internet of Things framework is proposed which is easily implementable, semantically distributive, and empowered by a machine learning model. The proposed system is equipped with a NodeRED dashboard which processes, visualizes, and stores the primary sensor data that are acquired through a public air quality sensor network, and further, the dashboard is integrated with a machine-learning model to obtain temporal and geo-spatial air quality predictions. ESP8266 NodeMCU is incorporated as a subscriber to the NodeRED dashboard via a message queuing telemetry transport broker to communicate quantitative air quality data or alarming emails to the end-users through the developed web and mobile applications. Therefore, the proposed system could become highly beneficial in empowering public engagement in air quality through an unoppressive, data-driven, and semantic framework.