Conditional Random Fields for High-Level Part Correlation Analysis in Images
A novel approach to model the semantic knowledge associated to objects detected in images is presented. The model is aimed at the classification of such objects according to contextual information combined to the extracted features. The system is based on Conditional Random Fields, a probabilistic graphical model used to model the conditional a-posteriori probability of the object classes, thus avoiding problems related to source modelling and features independence constraints. The novelty of the approach is in the addressing of the high-level, semantically rich objects interrelationships among image parts. This paper presents the application of the model to this new problem class and a first implementation of the system.
Unable to display preview. Download preview PDF.
- 1.Lafferty, J., McCallum, A., Pereira, F.: Conditional random fields: Probabilistic models for segmenting and labeling sequence data. In: Proc. 18th International Conf. on Machine Learning, pp. 282–289. Morgan Kaufmann, San Francisco (2001)Google Scholar
- 2.Kumar, S., Hebert, M.: Discriminative random fields: A discriminative framework for contextual interaction in classification. In: ICCV 2003. Proceedings of the 2003 IEEE International Conference on Computer Vision, vol. 2, pp. 1150–1157 (2003)Google Scholar
- 3.Quattoni, A., Collins, M., Darrell, T.: Conditional random fields for object recognition. In: Neural Information Processing Systems Vision (2004)Google Scholar
- 5.He, X., Zemel, R.S., Carreira-Perpinan, M.A.: Multiscale conditional random fields for image labeling. In: CVPR 2004. Computer Vision and Pattern Recognition, 2004. Proceedings of the 2004 IEEE Computer Society Conference on, vol. 2, pp. 695–702 (2004)Google Scholar