Markov Face Models

The spatial distribution of gray level intensities in an image can be naturally modeled using Markov Random Field (MRF) models. We develop and investigate the performance of face detection algorithms derived from MRF considerations. For enhanced detection, the MRF models are defined for every permutation of site indices in the image. We find the optimal permutation that provides maximum discriminatory power to identify faces from nonfaces. The methodology presented here is a generalization of the face detection algorithm in [1, 2] where a most discriminating Markov chain model was used. The MRF models successfully detect faces in a number of test images in real time. Key words and phrases: Markov Random Fields, face detection, maximum pseudolikelihood estimation, simulated annealing.

• Type: Conference
• Year: 2001
• Where: ICCV
• Authors: Sarat C. Dass, Anil K. Jain
• Keywords: Computer Vision, Face Detection Algorithm, Face Detection Algorithms, ICCV 2001, Markov Random Field, MRF Considerations, MRF Models