PRTools contents |
FACES
A = FACES(SUBJECTS,PHOTOS)
This creates a dataset A containing the desired photos (92*112 pixels) of the desired subjects (1 <= SUBJECTS <= 40). For each subject 10 photos are available (1 <= PHOTOS <= 10). SUBJECTS and PHOTOS may be vectors. The size of A is (LENGTH(SUBJECTS)*LENGTH(PHOTOS),92*112). The stored photos can be displayed by SHOW(A).
FACES first reads a datafile ORL and then converts it to a dataset. In case ORL is not available, it may be automatically downloaded and installed. If this fails an error is generated.
When using these images, please give credit to Olivetti Research Laboratory. A convenient reference is the face recognition work which uses some of these images
F. Samaria and A. Harter "Parameterisation of a stochastic model for human face identification",
2nd IEEE Workshop on Applications of Computer Vision
December 1994, Sarasota (Florida).
The paper is available via anonymous ftp from quince.cam-orl.co.uk and is stored in pub/users/fs/IEEE_workshop.ps.Z
If you have any question, please email Ferdinando Samaria: fs@cam-orl.co.uk
PRTools contents |