Bayesian classification and active learning using lp-priors. Application to image segmentation

Bayesian classification and active learning using lp-priors. Application to image segmentation. Ruiz, P., de la Blanca , N. P., Molina, R., & Katsaggelos, A. K. In 2014 22nd European Signal Processing Conference (EUSIPCO), pages 1183-1187, Sep., 2014.

Paper abstract bibtex

In this paper we utilize Bayesian modeling and inference to learn a softmax classification model which performs Supervised Classification and Active Learning. For p <; 1, lp-priors are used to impose sparsity on the adaptive parameters. Using variational inference, all model parameters are estimated and the posterior probabilities of the classes given the samples are calculated. A relationship between the prior model used and the independent Gaussian prior model is provided. The posterior probabilities are used to classify new samples and to define two Active Learning methods to improve classifier performance: Minimum Probability and Maximum Entropy. In the experimental section the proposed Bayesian framework is applied to Image Segmentation problems on both synthetic and real datasets, showing higher accuracy than state-of-the-art approaches.

@InProceedings{6952416,
  author = {P. Ruiz and N. P. {de la Blanca} and R. Molina and A. K. Katsaggelos},
  booktitle = {2014 22nd European Signal Processing Conference (EUSIPCO)},
  title = {Bayesian classification and active learning using lp-priors. Application to image segmentation},
  year = {2014},
  pages = {1183-1187},
  abstract = {In this paper we utilize Bayesian modeling and inference to learn a softmax classification model which performs Supervised Classification and Active Learning. For p <; 1, lp-priors are used to impose sparsity on the adaptive parameters. Using variational inference, all model parameters are estimated and the posterior probabilities of the classes given the samples are calculated. A relationship between the prior model used and the independent Gaussian prior model is provided. The posterior probabilities are used to classify new samples and to define two Active Learning methods to improve classifier performance: Minimum Probability and Maximum Entropy. In the experimental section the proposed Bayesian framework is applied to Image Segmentation problems on both synthetic and real datasets, showing higher accuracy than state-of-the-art approaches.},
  keywords = {belief networks;image classification;image segmentation;learning (artificial intelligence);maximum entropy methods;probability;Bayesian classification;active learning method;image segmentation problem;softmax classification model;supervised classification;variational inference;posterior probabilities;independent Gaussian prior model;minimum probability;maximum entropy;Bayes methods;Training;Support vector machines;Adaptation models;Image segmentation;Vectors;Entropy},
  issn = {2076-1465},
  month = {Sep.},
  url = {https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569924613.pdf},
}

Downloads: 0

{"_id":"BWo7Bjr9TPCCByN3e","bibbaseid":"ruiz-delablanca-molina-katsaggelos-bayesianclassificationandactivelearningusinglppriorsapplicationtoimagesegmentation-2014","authorIDs":[],"author_short":["Ruiz, P.","de la Blanca , N. P.","Molina, R.","Katsaggelos, A. K."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["P."],"propositions":[],"lastnames":["Ruiz"],"suffixes":[]},{"firstnames":["N.","P."],"propositions":["de la Blanca"],"lastnames":[],"suffixes":[]},{"firstnames":["R."],"propositions":[],"lastnames":["Molina"],"suffixes":[]},{"firstnames":["A.","K."],"propositions":[],"lastnames":["Katsaggelos"],"suffixes":[]}],"booktitle":"2014 22nd European Signal Processing Conference (EUSIPCO)","title":"Bayesian classification and active learning using lp-priors. Application to image segmentation","year":"2014","pages":"1183-1187","abstract":"In this paper we utilize Bayesian modeling and inference to learn a softmax classification model which performs Supervised Classification and Active Learning. For p <; 1, lp-priors are used to impose sparsity on the adaptive parameters. Using variational inference, all model parameters are estimated and the posterior probabilities of the classes given the samples are calculated. A relationship between the prior model used and the independent Gaussian prior model is provided. The posterior probabilities are used to classify new samples and to define two Active Learning methods to improve classifier performance: Minimum Probability and Maximum Entropy. In the experimental section the proposed Bayesian framework is applied to Image Segmentation problems on both synthetic and real datasets, showing higher accuracy than state-of-the-art approaches.","keywords":"belief networks;image classification;image segmentation;learning (artificial intelligence);maximum entropy methods;probability;Bayesian classification;active learning method;image segmentation problem;softmax classification model;supervised classification;variational inference;posterior probabilities;independent Gaussian prior model;minimum probability;maximum entropy;Bayes methods;Training;Support vector machines;Adaptation models;Image segmentation;Vectors;Entropy","issn":"2076-1465","month":"Sep.","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569924613.pdf","bibtex":"@InProceedings{6952416,\n author = {P. Ruiz and N. P. {de la Blanca} and R. Molina and A. K. Katsaggelos},\n booktitle = {2014 22nd European Signal Processing Conference (EUSIPCO)},\n title = {Bayesian classification and active learning using lp-priors. Application to image segmentation},\n year = {2014},\n pages = {1183-1187},\n abstract = {In this paper we utilize Bayesian modeling and inference to learn a softmax classification model which performs Supervised Classification and Active Learning. For p <; 1, lp-priors are used to impose sparsity on the adaptive parameters. Using variational inference, all model parameters are estimated and the posterior probabilities of the classes given the samples are calculated. A relationship between the prior model used and the independent Gaussian prior model is provided. The posterior probabilities are used to classify new samples and to define two Active Learning methods to improve classifier performance: Minimum Probability and Maximum Entropy. In the experimental section the proposed Bayesian framework is applied to Image Segmentation problems on both synthetic and real datasets, showing higher accuracy than state-of-the-art approaches.},\n keywords = {belief networks;image classification;image segmentation;learning (artificial intelligence);maximum entropy methods;probability;Bayesian classification;active learning method;image segmentation problem;softmax classification model;supervised classification;variational inference;posterior probabilities;independent Gaussian prior model;minimum probability;maximum entropy;Bayes methods;Training;Support vector machines;Adaptation models;Image segmentation;Vectors;Entropy},\n issn = {2076-1465},\n month = {Sep.},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569924613.pdf},\n}\n\n","author_short":["Ruiz, P.","de la Blanca , N. P.","Molina, R.","Katsaggelos, A. K."],"key":"6952416","id":"6952416","bibbaseid":"ruiz-delablanca-molina-katsaggelos-bayesianclassificationandactivelearningusinglppriorsapplicationtoimagesegmentation-2014","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569924613.pdf"},"keyword":["belief networks;image classification;image segmentation;learning (artificial intelligence);maximum entropy methods;probability;Bayesian classification;active learning method;image segmentation problem;softmax classification model;supervised classification;variational inference;posterior probabilities;independent Gaussian prior model;minimum probability;maximum entropy;Bayes methods;Training;Support vector machines;Adaptation models;Image segmentation;Vectors;Entropy"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2014url.bib","creationDate":"2021-02-13T17:43:41.674Z","downloads":0,"keywords":["belief networks;image classification;image segmentation;learning (artificial intelligence);maximum entropy methods;probability;bayesian classification;active learning method;image segmentation problem;softmax classification model;supervised classification;variational inference;posterior probabilities;independent gaussian prior model;minimum probability;maximum entropy;bayes methods;training;support vector machines;adaptation models;image segmentation;vectors;entropy"],"search_terms":["bayesian","classification","active","learning","using","priors","application","image","segmentation","ruiz","de la blanca ","molina","katsaggelos"],"title":"Bayesian classification and active learning using lp-priors. Application to image segmentation","year":2014,"dataSources":["A2ezyFL6GG6na7bbs","oZFG3eQZPXnykPgnE","ya2CyA73rpZseyrZ8","KTWAakbPXLGfYseXn","ePKPjG8C6yvpk4mEK","D8k2SxfC5dKNRFgro","7Dwzbxq93HWrJEhT6","qhF8zxmGcJfvtdeAg","fvDEHD49E2ZRwE3fb","H7crv8NWhZup4d4by","DHqokWsryttGh7pJE","vRJd4wNg9HpoZSMHD","sYxQ6pxFgA59JRhxi","w2WahSbYrbcCKBDsC","XasdXLL99y5rygCmq","3gkSihZQRfAD2KBo3","t5XMbyZbtPBo4wBGS","bEpHM2CtrwW2qE8FP","teJzFLHexaz5AQW5z"]}