AI MultiMedia Lab.

The rapid increase in multimedia data traffic, now exceeding 70% of total Internet traffic, underscores the expanding demand for on-demand video, private broadcasts, and the rising prominence of immersive formats like virtual reality (VR) and augmented reality (AR). These new formats, however, require significantly higher computing power and network bandwidth to deliver the immersive and high-quality experience users expect.

Despite their popularity, there’s limited research on how viewers perceive VR/AR content quality, especially when it’s delivered over wired or wireless networks. This perception is crucial to enhancing the overall multimedia experience. AIMM Labs addresses this gap by developing AI-powered multimedia computing systems focused on delivering VR/AR data more efficiently across networks. Our goal is to ensure high-quality, seamless multimedia experiences by leveraging AI to optimize resource allocation, processing, and delivery of rich multimedia content.

Research topics

Here are some key research areas:

Semantic-Aware View Prediction for On-Demand VR Video Streaming Systems:

Focuses on predicting user views in VR to enhance the streaming quality based on semantic understanding.

Semantic-Aware View Prediction for Live VR Video Streaming Systems:

Concentrates on real-time view prediction in VR environments to improve live streaming experiences through semantic insights.

3D Point Cloud Data Streaming for AR Applications and its Quality of Experience (QoE):

Investigates efficient streaming and enhancing the QoE in augmented reality applications by managing 3D point cloud data.

Adaptive 3D Point Cloud Data Dissemination for Autonomous Driving Systems:

Involves the dissemination of 3D point cloud data to support autonomous driving, focusing on rate adaptation algorithm