Monthly Archives: November 2008

Video Summaries for YouTube Videos?

Applying old things to new platforms has become common in recent times, here’s my contribution. I recently developed a video summary tool based on FFMPEG and Lire for a friend … just to test if common approaches are usable in a specific domain. Video summarization – especially of small videos – is a rather easy thing. You just need to find a number of frames with maximized pairwise difference, to cover a maximized visual range of the video. I applied my tool on YouTube and got the following summaires for the “hippo bathing” video:

Based on the CEDD descriptor the most important keyframe is really chosen well – just watch the video to know what I mean :)

With the auto color correlogram feature the dog is not explicitely part of the picture. However the first frame chosen (the big one) gives a good impression on the “bathing” part.

With the Gabor texture feature the dog gets prominent in the first place. Noite that the result is quite the same as the result kwith the Tamura texture feature not shown here.

With the most simple feature (RGB color histogram with L2 distance) the summary also looks appealing. There is a frame featuring the dog, one showing the whole scene and one for the hippo.

All in all I think the results are quite appealing. The runtime of my implementation is a fraction of actual video play time. Perhaps I’ll find some time to present the whole thing tomorrow at the barcamp ;)

Report from the Open Source Track @ ACM MM

Within the open source track of the ACM Multimedia conference the SIGMM  (Special Interest Group on Multimedia of the ACM) wants to provide a platform on discussing tools and code free to use in research as well as industry project. As I already blogged I had a presentation on Lire there. However all other contributors were nearly as interesting, if not even more ;)

Marco Lohse presented a Network-Integrated Multimedia Middleware (NMM), which was developed at Saarland University and is now available with dual licensing. He is also CEO of the company selling the non GPL license. The software is impressive and his demo was great. No surprise he won the prize of 700 $. (http://www.motama.com/nmm.html)

Yannick Alluse presented GpuCV, an extension of OpenCV towards CUDA and shader based computing. Important thing is that the library decides whether CPU, Shader or the GPU will be used for processing. (https://picoforge.int-evry.fr/cgi-bin/twiki/view/Gpucv/Web/)

Andreas Berger showed a framework for DVB-* transmission. It allows for inexpensive setup of for instance a DVB-T or DVB-H broadcasting service in labs or research projects. As its very generic one can compose the DVB stream fully free, so MHP, IP Data, EPG and so on are possible. (http://sourceforge.net/projects/fatcaps/)

Jose San Pedro Wandelmer from the University of Sheffield presented his project FOBS, which is actually a wrapper for ffmpeg for C++ and JMF. I already knew FOBS as I have used it several times, but meeting Jose and talking to him was great. (http://fobs.sourceforge.net/)

Next week: Barcamp Graz, AT

At the end of next week a barcamp in Graz is taking place.  As always there is no main topic under the assumption that one will be emerging from the talks of the people contributing. Currently there is a temporary location; however a bigger one is still searched for. So if you know one?!?

I already scheduled my trip to Graz and I will definitely present something there. My current favorite idea is a talk about the capabilities of content based multimedia search doing some disillusionment (if this is a real word).

Links:

ACM Multimedia 2010 in Florence, IT

Although my boss Laszlo Böszörmenyi did a great job on presenting our bid for hosting ACM Multimedia 2010, we didn’t stand a chance against Florenze. Alberto del Bimbo and his group will host the ACM Multimedia in 2010. Although we are slightly sad because we thought we will get it for sure (we didn’t know our competitors), we are now quite relieved that (i) another one is doing the organizational work and (ii) we had a good opportunity to present our department & research groups within the ACM MM community. We’ll definitely be there in Florence and help the group of Alberto del Bimbo as good as we can.