Constraining probabilistic relaxation with symbolic attributes

Hatef, M.; Kittler, Josef

doi:10.1007/3-540-60268-2_394

Search citation statements

Order By: Relevance

Paper Sections

Select...

Experimental Comparison Of Classifier Combination Rules: Han1

The Classification Scheme1

Citation Types

Supporting

Mentioning

Contrasting

Year Published

1996

1998

Publication Types

Select...

Article1

Other1

Relationship

Self Cite0

Independent2

Authors

Journals

Cited by 2 publications

(2 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Four types of classifiers are first applied to perform the classification individually. We used structural [38], Gaussian, Neural Network, and Hidden Markov Model classifiers [40].…”

Section: Experimental Comparison Of Classifier Combination Rules: Hanmentioning

confidence: 99%

“…Otherwise, the sample will be checked against all prototypes in the code-Group to find the closest candidate(s) to the sample. First, the probabilistic relaxation algorithm [38] is used to find the correspondence between the primitives in the sample and those in the prototype. Then, a distance measure is used to quantify the similarity between the sample and each candidate.…”

Section: The Classification Schemementioning

confidence: 99%

See 1 more Smart Citation

On combining classifiers

Kittler

Hatef

Duin

et al. 1998

IEEE Trans. Pattern Anal. Machine Intell.

4,501

1,892

View full text Add to dashboard Cite

We develop a common theoretical framework for combining classifiers which use distinct pattern representations and show that many existing schemes can be considered as special cases of compound classification where all the pattern representations are used jointly to make a decision. An experimental comparison of various classifier combination schemes demonstrates that the combination rule developed under the most restrictive assumptions-the sum rule-outperforms other classifier combinations schemes. A sensitivity analysis of the various schemes to estimation errors is carried out to show that this finding can be justified theoretically.

show abstract

“…Four types of classifiers are first applied to perform the classification individually. We used structural [38], Gaussian, Neural Network, and Hidden Markov Model classifiers [40].…”

Section: Experimental Comparison Of Classifier Combination Rules: Hanmentioning

confidence: 99%

Section: The Classification Schemementioning

confidence: 99%