This paper presents a transformer-based machine learning model that aligns gaze trajectory with corresponding text and decodes the probability of unknown words within the text in real time with high precision. 
FlowRing enables both quick, discreet in-air microgestures (left) and precise on-surface interactions (right), allowing for always-available, ad-hoc interactions. Such a peripheral could control wearable devices quickly like changing earbud volume while on-the-go or be used to control an immersive desktop environment in a mixed reality headset.
FaceOri tracks user’s face orientation towards the device with acoustic ranging using microphones in an earphone. 
AcousLink enables ad-hoc calibration-free unltrasonic ranging via bluetooth. 
Our method locates the content the user is reading in real-time through gaze, and inputs the gaze data and text data to the transform-based model to detect unknown words. 
The Smart Frisbee project aims to address the limitations of traditional Ultimate Frisbee games by utilizing Bluetooth and IMU sensors in a smart frisbee and wristbands to collect and analyze data on throws, catches, and player performance. This data will be used to provide automated tracking, feedback, and analysis capabilities, enabling players to identify areas for improvement and enhance their gameplay experience.
