Pervasive Artificial Intelligence Research (PAIR) Labs

Drones and Computer Vision

Pervasive Artificial Intelligence Research (PAIR) Labs

Drones and Computer Vision

NCTU – Prof. Wen-Hsiao Peng (Prof. Hsueh-Ming Hang)

zjvjZ7nYJS7sGvhWbkmLxR5Tu5ITFIH1D4rXFvNS

Principal Investigator: Prof. Wen-Hsiao Peng (Prof. Hsueh-Ming Hang)

Research：

Video Compression based on Generative Model

Core Technologies

Image and video compression using deep learning technololgy. Currently, it is still under development but we believe it has tremendous potential. If it outperforms the certain image/video compression standards, it could possibly become international video standard after five years. And it could use as a compression framework for internal network system usage in the near future.

Applications

Smart home applications
1. multimedia server
2. Set-up box
3. smart TV
4. smart home security camera

Technology Transfe

Learning-based image compression

Technologies with Potential for Technology Transfer or Industry-Academia Collaboration

Learning-base video compression
1. Video preprocessing
2. Video coding for machine
3. RL-based video encoder control

Proposal for Industry-Academia Collaboration

Use deep learning and latest video compression techniques to improve the image transmission for private clouds / personal multimedia servers.
(e.g.
1. Maintain acceptable image quality under ultra low bit rate condition.
2. Reduce image artifact under multiple video/image format conversions (for streaming usage)
3. The compression framework can run on edge devices, such as phone, tablets, etc.
)