KEYWORD |
Use of h264/h265 motion vectors as objects motion estimationy
Thesis in external company
keywords CODEC, H264, H265
Reference persons BARTOLOMEO MONTRUCCHIO
External reference persons Enrico Busto enrico.busto@add-for.com
Research Groups GR-09 - GRAphics and INtelligent Systems - GRAINS
Thesis type RESEARCH
Description AVC and HEVC two of the most used codecs in video industry internally uses intra and extra frame motion vectors in order to achieve a better compression.
A very recurring task in computer vision is motion detection and a lot of solutions are already been given. Most of these solutions are very heavy on the underling hardware. Hence it will be very interesting to understand if (and how) we can exploit these information that are already in the video stream to achieve similar results.
SEE ALSO:
[1]https://pdfs.semanticscholar.org/98ca/f8325abb40aa6bbddd0e7f5d3a6c366d03e6.pdf
[2]https://pdfs.semanticscholar.org/06ac/28ba7c9d5813564f1421c4600503aa82795d.pdf
[3]http://image.ntua.gr/iva/files/papers/Robust%20optical%20flow%20estimation%20in%20MPEG%20sequences%20-%20K.%20Rapantzikos.%20M.%20Zervakis2005.pdf
[4]https://www.semanticscholar.org/paper/Approximating-optical-flow-within-the-MPEG-2-domain-Coimbra-Davies/902d13657cf56287624ddf1cd8c1bb18b0f07ddd
Required skills Low level programming, C/C++, Python programming, basic knowledge of Machine Learning.
Deadline 01/06/2020
PROPONI LA TUA CANDIDATURA