[mythtv] PATCH: Commercial detection/skip using blank frames
Vincent AE Scott
mythtv at codex.dyndns.tv
Fri Feb 28 03:47:52 EST 2003
Brent Norris(brent at biglinux.tccw.wku.edu)@Thu, Feb 27, 2003 at 09:12:04PM -0600:
> The problem with this is when you record something like basketball games
> or football games or NASCAR races. Often at the end of these events
> there are periods where several timeouts occur where the station goes to
> commercial, but in between these breaks there is often extremely
> important things happening. If you use the statistical model for
> something like this it will either skip an important play or not
> recognize a block of commercials. I remember watching a basketball game
> one time where there were 4 30 second commercial breaks with about 10-15
> seconds of game inbetween them.
Yes, for certain programs i can see the analysis would prove
problematic. which is why associating models with stations/programs
would be very usefull. The heuristics for one program's advert
detection might not be the same as anothers. If the system could
remember that live sports games, maybe subject to short commercial
breaks, or other non typical segmentation, then that too could be used
to try and spot ads.
maybe, if you know that an avergae hours worth of TV has 20 mins
of adverts crammed into it you could use that knowledge to determine teh
likelihood of an ad occuring. i.e. in the last 15 minutes of program,
you might know with a high degree of certainty that there will be
another 5 mins of ads. Possibly helping to provide higher weightings
associated with possible ads that have been detected.
> On Thu, 2003-02-27 at 20:49, Vincent AE Scott wrote:
> > I've been following this thread with much anticipation, altho i don't
> > have an actual fully working mthytv setup, this for me is certainly an
> > awesome feature.
> > I like this idea of tagging the ads on the fly, for later post recording
> > analysis, that got me thinking:
> <huge snip>
> </huge snip>
> > As i said at the start, this is a fascinating topic, especially given
> > what the eventual outcomes might be. In my enthusiasm i hope i haven't
> > over looked something, or that this has already been better explained by
> > someone else previously. Better to have ideas re-iterated than
> > forgotten.
> > ttfn,
> > -v
> Brent Norris <brent at linux.wku.edu>
> mythtv-dev mailing list
> mythtv-dev at snowman.net
keys: http://codex.net/pgp/gpg.asc http://codex.net/pgp/pgp.asc
We don't really understand it,
so we'll give it to the programmers.
More information about the mythtv-dev