Computer Science related thesis

Placebo

Hi, not sure where this belongs but think il get most relevant response here.
Currently doing Masters in MSC, ive a Bsc in computer science, ive to write a thesis for my course on a topic of my choice,
Topic should be multimedia related, so i put forward a few ideas such as face recognition etc but ive been told its too tech, so im really stuck for ideas here.
If anyone could throw something at me, it would be very helpful.
I was contemplating something on web 2.0 but it must stretch to 12k words
Ajax has already been done so i cant really branch into that.

Giblet

Might not be helpful seeing as you were told image recognition was too tech, but there is a lot of research going into terahertz waves, which are basically like x-rays which aren't harmful. The images pulled from these machines show a semi-transparent image, so in an airport you could use it to scan people for objects and do recognition on those objects. Finding a gun from different angles etc etc.

OMcGovern

Face recognition can be tricky alright, teaching neural nets to recognise faces at various sizes.

A couple of vaguely multi-media areas to consider
- vehicle number plate recognition
- RFID tags, their potential uses for anti-shop lifting, tracking items/people, security controlled vehicle barriers, etc...
There's cheap developer kits out there...
http://www.trossenrobotics.com/store/c/2785-Starter-Kits.aspx?a=phidgetsusaRFID

- Google maps, maybe some GPS tracking app linked to it, or allow users to upload pics/videos and link them to their location.
Or map out all public transport links in Dublin ( bus, rail, luas ), and write an application to get you from point A to point B. ( With or without proper scheduling to minimise waiting times etc... )

- Natural language processing, trying to parse web pages for meaning rather than doing simple text searching on them.
- Idle cpu time, as PC's get faster, is there anything they can do for businesses in their spare time.. much like the SETI screensaver approach.
- Use of personal webcams as next phase of "neighbourhood watch" schemes
Security versus privacy. Could be linked with motion detection if you
wanted to highlight areas of interest in a long recording... maybe detect
vehicles, people etc...

GoneShootin

When I was doing my thesis I originally played around with the idea of Software Watermarking. Twas interesting, but not my cup of tea.

hurlerman

eLearning - how multimedia is used to teach... say with second life - podcasts - video

Maybe how it can advance in the future - it's very basic at the moment

Or how it can multimedia can be used to help people with disabilities learn

Tom Dunne

Placebo wrote: »

Hi, not sure where this belongs but think il get most relevant response here.
Currently doing Masters in MSC, ive a Bsc in computer science, ive to write a thesis for my course on a topic of my choice,
Topic should be multimedia related, so i put forward a few ideas such as face recognition etc but ive been told its too tech, so im really stuck for ideas here.
If anyone could throw something at me, it would be very helpful.
I was contemplating something on web 2.0 but it must stretch to 12k words
Ajax has already been done so i cant really branch into that.

RFID is certainly a hot-topic technology, but it's hardly multimedia (God, I hate that word).

What you need to do is pick a topic and give it an edge that is worthy of an MSc thesis.

For example, the likes of YouTube - take it one step further to mobiles, investigating the technology needed or develop a framework. Taking another example, what about delivering video-based advertisements to mobile phones in exchange for free content?

I think the whole e-learning thing has been done to death, TBH. You would want a very good slant to get it noticed.

What about a voice-controlled home entertainment system? Or a program to alllow a person access their mp3's on their home PC from anywhere on the internet (on a 3G mobile phone, perhaps?).

</brain dump>

Placebo

hey guys , thanks alot for those suggestions,
im not sure whats really expected, quite vague requirements.Judging from the past thesis, i dont think anyone has tackled a unique subject at all !
Id like to do something on google map api, even do a RIA for it but i dont think thats required at all, although i could start off talking about the api and capabilities il need my own idea to implement, bus's sounds great !

thanks again