Interactive AR/VR Systems with Deep Learning

Подробнее о видео

Recent advancement in Graphics Processing Units (GPU) made augmented and virtual reality systems commonplace and even available in low cost platforms. Augmented and Virtual reality systems have tremendous potential to re-shape our experience with digital media in terms of various degrees of immersion and interaction. Advancement in GPUs along with artificial intelligence also redefined interaction with electronic media in the form of eye gaze, gesture and body movement tracking in real-time with low cost imaging devices.

In the context of interactive AR/VR systems, this talk will present our research on integrating deep learning models with augmented and virtual reality environments for both traditional use cases of process modelling in smart manufacturing set up to novel applications of rehabilitation of people with disabilities, pilot training and enforcing social distance measures in office spaces. In particular, I shall present the continuum of displays, alternative input modalities like eye gaze tracking, gesture and speech recognition and basic theories of deep learning models like Convolutional Neural Networks. The integration of eye gaze tracking, gesture recognition and deep learning with AR/VR systems will be demonstrated through case studies.