Things have become a little more complicated since my last post, I have been looking at the technicalities of developing and producing my model and found this is more complicated than first imagined. There is two parts to the idea, 1. The 3d model. 2. Facial recognition system. each has an element of technical difficulty to overcome, however the main problem is getting the two to interact. I recently had a conversation with a games design tutor who very kindly gave me some extremely useful advice on 3d model expressions and the lack of realism of these expressions. This has the potential to increase the technical difficulty with regard to the models expressive detail, further more this appears to be a problem throughout the 3d modelling world.
The facial recognition system is also proving difficult to match to the proposal, many FR systems are based on a contrast comparison model and as such can operate at a low camera resolution, which is fine for most uses. However, I need the system to pick up small changes of each quarter of the face and recognise these changes.
Further investigation into the technical realities of this project is needed.
On a positive note I have found a neat pice of software that allows you to create models easily, however you can only use them if you purchase the license. see link below
http://www.genemation.com/New_Pages/gentools.cfm
It also uses a tracking system, so I think an email is needed.
Monday, 26 November 2007
Sunday, 4 November 2007
Ideas for Hybrid unit
This week we had to present our idea for the hybrid unit, the idea I have is based around the using digital technology to assist face recognition and social communication in people with the ASD condition. My objective is to develop a visual program that can be used to facilitate the recognition of facial expressions and social nuances.
The program will start at a basic level and progress in levels to more complex expressions and meanings. The problem is getting the software to function accurately, and to generate the 3d facial model. Research into face recognition aids has mainly been video based using a selection process and using actors. My idea will be to progress this further into a 3d model form. I have to figure out exactly how as yet.
If anyone can offer possible technical advice it will be much appreciated.
The program will start at a basic level and progress in levels to more complex expressions and meanings. The problem is getting the software to function accurately, and to generate the 3d facial model. Research into face recognition aids has mainly been video based using a selection process and using actors. My idea will be to progress this further into a 3d model form. I have to figure out exactly how as yet.
If anyone can offer possible technical advice it will be much appreciated.
Subscribe to:
Posts (Atom)
