3D face recognition
Context
Face recognition is a very popular research area in the field of computer vision. Due to the nature of the problem it addresses questions to researchers from various disciplines such as image processing, machine learning, data mining or psychology. We are interested in face recognition methodologies that make use of physiological characteristics in order to accurately identify an individual. In general terms, the face recognition problem can be formulated as follows: given one or more images containing faces of a person, identify that person using a stored database of images. The first biometrical systems for face recognition made use of 2D images serving as knowledge base for feature extraction in order to solve the recognition problem. There is a vast literature on 2D face recognition methods and many different implementations of algorithms, which can serve for the understanding of the problem.
Goal
The first task is to explore the state of the art methods for 2D face recognition and to investigate new feature extraction approaches in order to improve the recognition rate of state of the art methods. Particularly we aim to investigate a new Non Negative Matrix Factorization method based on genetic strategy optimization for feature extraction. Open challenged in this area include:
- Recognition from non-frontal facial images
- Emotion recognition from face images
- Gender recognition
to mention only few of them. The candidate is required to investigate solutions for one of the three particular problems mentioned above. However 2D face recognition has important limitations when external factors such as illumination or head pose vary. The actual scientific trend to solve these limitations is to use 3D images for face recognition. The second task in this proposal is to build a biometric system for 3D face recognition. Within this project, the candidate disposes of a 3D camera, which allows the candidate to build a database of 3D images for training and tests. The candidate should investigate state of the art approaches proposed for 2D face recognition but also, as further research work, to propose his/her ideas for 3D face recognition in one of the problems mentioned above.
Contacts
- Supervisor: Cosmin Lazar
- Promoter: Ann Nowé
