Combining belief networks and neural networks for scene segmentation

Xiaojuan Feng, C.K.I. Williams, S.N. Felderhof

Research output: Contribution to journalArticlepeer-review

Abstract / Description of output

We are concerned with the problem of image segmentation, in which each pixel is assigned to one of a predefined finite number of labels. In Bayesian image analysis, this requires fusing together local predictions for the class labels with a prior model of label images. Following the work of Bouman and Shapiro (1994), we consider the use of tree-structured belief networks (TSBNs) as prior models. The parameters in the TSBN are trained using a maximum-likelihood objective function with the EM algorithm and the resulting model is evaluated by calculating how efficiently it codes label images. A number of authors have used Gaussian mixture models to connect the label field to the image data. We compare this approach to the scaled-likelihood method of Smyth (1994) and Morgan and Bourlard (1995), where local predictions of pixel classification from neural networks are fused with the TSBN prior. Our results show a higher performance is obtained with the neural networks. We evaluate the classification results obtained and emphasize not only the maximum a posteriori segmentation, but also the uncertainty, as evidenced e.g., by the pixelwise posterior marginal entropies. We also investigate the use of conditional maximum-likelihood training for the TSBN and find that this gives rise to improved classification performance over the ML-trained TSBN
Original languageEnglish
Pages (from-to)467-483
Number of pages17
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Issue number4
Publication statusPublished - 1 Apr 2002

Keywords / Materials (for Non-textual outputs)

  • belief networks
  • feature extraction
  • image coding
  • image segmentation
  • learning (artificial intelligence)
  • maximum likelihood estimation
  • multilayer perceptrons
  • probability
  • Bayesian image analysis
  • EM algorithm
  • Gaussian mixture model
  • class labels
  • classification performance
  • conditional maximum-likelihood training
  • expectation-maximization
  • label images
  • local predictions
  • maximum a posteriori segmentation
  • maximum-likelihood objective function
  • neural networks
  • pixel classification
  • pixelwise posterior marginal entropies
  • scaled-likelihood method
  • scene segmentation
  • tree-structured belief networks
  • uncertainty
  • Layout
  • Neural networks


Dive into the research topics of 'Combining belief networks and neural networks for scene segmentation'. Together they form a unique fingerprint.

Cite this