A decision-theoretic generalization of on-line learning and an application to boosting
Journal of Computer and System Sciences - Special issue: 26th annual ACM symposium on the theory of computing & STOC'94, May 23–25, 1994, and second annual Europe an conference on computational learning theory (EuroCOLT'95), March 13–15, 1995
IEEE Transactions on Pattern Analysis and Machine Intelligence
Recognizing Action Units for Facial Expression Analysis
IEEE Transactions on Pattern Analysis and Machine Intelligence
Input Decimation Ensembles: Decorrelation through Dimensionality Reduction
MCS '01 Proceedings of the Second International Workshop on Multiple Classifier Systems
Comprehensive Database for Facial Expression Analysis
FG '00 Proceedings of the Fourth IEEE International Conference on Automatic Face and Gesture Recognition 2000
Efficient Feature Selection via Analysis of Relevance and Redundancy
The Journal of Machine Learning Research
Fully Automatic Facial Action Recognition in Spontaneous Behavior
FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
Feature Ranking Ensembles for Facial Action Unit Classification
ANNPR '08 Proceedings of the 3rd IAPR workshop on Artificial Neural Networks in Pattern Recognition
Solving multiclass learning problems via error-correcting output codes
Journal of Artificial Intelligence Research
Stopping criteria for ensemble-based feature selection
MCS'07 Proceedings of the 7th international conference on Multiple classifier systems
Combining feature subsets in feature selection
MCS'05 Proceedings of the 6th international conference on Multiple Classifier Systems
Accuracy/Diversity and Ensemble MLP Classifier Design
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
Facial action unit (au) classification is an approach to face expression recognition that decouples the recognition of expression from individual actions. In this paper, upper face ausare classified using an ensemble of MLP (Multi-layer perceptron) base classifiers with feature ranking based on PCA components. This approach is compared experimentally with other popular feature-ranking methods applied to Gabor features. Experimental results on Cohn-Kanade database demonstrate that the MLP ensemble is relatively insensitive to the feature-ranking method but optimized PCA features achieve lowest error rate. When posed as a multi-class problem using Error-Correcting-Output-Coding (ECOC), error rates are comparable to two-class problems (one-versus-rest) when the number of features and base classifier are optimized.