Parallel media processing

Using multi-core processors can reduce the latency of video encoding and transcoding. How close can you get to a linear speed increase? Is it possible no matter which encoding parameter you choose for the video?

For networked multimedia applications, such as video conferencing, live video broadcasting, the end-to-end playback latency is a very stringent QoS measure. It is required to constant and has very a limited budget. Among different components contributing to the end-to-end delay, such as network transmission, scheduling and the media processing time, the encoding time is significantly important. The media compression process is computation demanding not only because of the large amount of the media data, but also because many encoding options need to be evaluated for improving the compression efficiency.

The objective of this thesis is how to speed up the encoding process by parallel computing on multiprocessors.

Prerequisites: English, excellent programming skills, C or C++, Fundamentals of multimedia, Parallel computing, Computer architecture

For more information please contact Pengpeng Ni
Personal tools