On a New Class of Bounds on Bayes Risk in Multihypothesis Pattern Recognition

Devijver, Pierre A.

doi:10.1109/t-c.1974.223779

Cited by 61 publications

(32 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…I missed the 1973 ICPR in Washington DC where Pierre first introduced his ideas about the relationship of the Bayes Risk to the Mean Square Error [2]. He published an article on error bounds the following year in the IEEE Transactions on Computers (where the best papers on pattern recognition were published before PAMI came along) [3]. So even before I ever met him I had studied some of his work.…”

Section: Pierre Devijvermentioning

confidence: 99%

Estimation, Learning, and Adaptation: Systems That Improve with Use

Nagy¹

2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The accuracy of automated classification (labeling) of single patterns, especially printed, hand-printed, or handwritten characters, has leveled off. Further gains in accuracy require classifying sequences of patterns. Linguistic context, already widely used, relies on 1-D lexical and syntactic constraints. Style-constrained classification exploits the shape-similarity of sets of same-source (isogenous) characters of either the same or different classes. For understanding tables and forms, 2-D structural and relational constraints are necessary. Applications of pattern recognition that do not exceed the limits of human senses and cognition can benefit from green interaction wherein operator corrections are recycled to the classifier.

show abstract

Section: Pierre Devijvermentioning

confidence: 99%

Estimation, Learning, and Adaptation: Systems That Improve with Use

Nagy¹

2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…We therefore need an entropy measure that weights the lower-valued coefficients less than the higher-valued ones. This is a property of a Bayes entropy function (Devijver, 1974), defined as…”

Section: Entropy Of a Normalized Spectrummentioning

confidence: 99%

A Bayes-spectral-entropy-based measure of camera focus using a discrete cosine transform

Kristan

Perš

Perše

et al. 2006

Pattern Recognition Letters

101

View full text Add to dashboard Cite

In this paper we present a novel measure of camera focus based on the Bayes spectral entropy of an image spectrum. In order to estimate the degree of focus, the image is divided into non-overlapping subimages of 8 by 8 pixels. Next, sharpness values are calculated separately for each sub-image and their mean is taken as a measure of the overall focus. The sub-image spectra are obtained by an 8 × 8 discrete cosine transform (DCT). Comparisons were made against four well-known measures that were chosen as reference, on images captured with a standard visible-light camera and a thermal camera. The proposed measure outperformed the reference measures by exhibiting a wider working range and a smaller failure rate. To assess its robustness to noise, additional tests were conducted with noisy images.

show abstract

“…A number of upper and lower bounds have been proposed in the literature [51][52][53][54][55][56][57][58][59][60][61][62][63].…”

Section: Bounds On the Probability Of Error Of Optimum Receiversmentioning

confidence: 99%

“…A tighter bound on the probability of error than the Bhattacharyya bound is in terms of the equivocation function [55,53]. Devijver [56] introduced another bound in terms of the so called Bayesian distance. This bound is known to be tighter than both the Bhattacharyya bound and the equivocation bound.…”

Section: Bounds On the Probability Of Error Of Optimum Receiversmentioning

confidence: 99%

“…A number of upper and lower bounds on the MPOE of optimum receivers have been proposed in the literature [51][52][53][54][55][56][57][58][59][60][61][62][63]. Because of the indirect relationship between the probability of error and the class of Ali-Silvey distance measures (or the f-divergence) [31], a number of the bounds available are expressed in terms of these distance measures.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation