System Information Sciences
Real-time Computing Systems B17
Real-time Edge Computing for Unprecedented-scale Data
Systems in which processing tasks must be completed within appropriate time constraints are referred to as real-time systems. Real-time performance is critical in application areas where systems interact closely with the physical world, such as robotics and human interfaces. Moreover, in today's world where massive amounts of data are generated continuously across the globe, real-time edge computing, which extracts only useful information in order to stay within the limits of communication bandwidth and storage capacity, has become increasingly important.
Our laboratory conducts research on a wide range of sensory information processing based on high-speed real-time data processing, along with its applications. We focus especially on video and 3D data, with an emphasis on the use of high-frame-rate video. Our work addresses challenges in application fields such as robotics, advanced scientific measurement, video media, human-computer interaction, and virtual/augmented reality, including the following:
- Real-time information extraction from high-frame-rate video and dynamic image capture control
- High-speed multidimensional data acquisition using active light source control
- Edge data processing for low-latency video display and its application to visualization and interaction
-
Implementation of a low-latency projector with an embedded edge computing engine
-
Calibration-free low-latency projection mapping