I want to incorporate optimal feature selection in high dimensional discriminant analysis in order to effectively improve the performance of face recognition algorithm. My specific question would be, is it possible? and how will I address the common issue of face recognition?
I want to have clear directions of my study. You can also suggest recent research related to this.
As you must be knowing HDDA is one of the efficient algorithms for object detection.
If you are wondering whether you can use it for face recognition or not that will mainly depend on your own implementation that is, how well do you define your "face" class (or you detect multiple objects of face simultaneously and deduce recognition by considering cumulative result).
On the other hand I would suggest that before executing your face recognition do provide some filters to eliminate noise and solve illumination and brightness issues.
The outcome will also depend on the camera that you use. But in a nutshell, Yes you can give it a try.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Algorithms