Strategies for combining classifiers employing shared and distinct pattern representations

Kittler, Josef; Hojjatoleslami, Ali; Windeatt, Terry

doi:10.1016/s0167-8655(97)00095-0

Cited by 34 publications

(15 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, the intrinsic difficulty of the partially unsupervised classification problem results in classifiers that are less reliable and less accurate than the corresponding supervised ones, especially for complex data sets. Therefore, by taking into account that, in general, ensembles of classifiers are more accurate and more robust than the individual classifiers that make them up [5], we expect that a multipleclassifier approach may increase the reliability and the accuracy of the global classification system. A further step aimed at improving the performance of the system consists in implementing each partially unsupervised classification algorithm of the ensemble in the framework of a cascade-classifier approach, thus exploiting also the temporal correlation between the multitemporal images in the updating process.…”

Section: B System Architecturementioning

confidence: 99%

“…In the literature, several different strategies for defining a classifier ensemble have been proposed [5], [14]- [17]. From a theoretical viewpoint, necessary and sufficient conditions for an ensemble of classifiers to be more accurate than any of its individual members are that the classifiers should be accurate and different [18].…”

Section: A Strategy For Generating Ensembles Of Partially Unsupermentioning

confidence: 99%

“…In particular, we consider two simple and widely used combination procedures: Majority Voting and Combination by Bayesian Average [5]. Both procedures exhibit the common peculiarity of requiring no prior training to carry out the combination process.…”

Section: Multiple Cascade Classifier Architecture: Unsupervised Comentioning

confidence: 99%

See 2 more Smart Citations

A multiple-cascade-classifier system for a robust and partially unsupervised updating of land-cover maps

Bruzzone

Cossu

2002

IEEE Trans. Geosci. Remote Sensing

112

View full text Add to dashboard Cite

show abstract

Section: B System Architecturementioning

confidence: 99%

Section: A Strategy For Generating Ensembles Of Partially Unsupermentioning

confidence: 99%

See 1 more Smart Citation

A multiple-cascade-classifier system for a robust and partially unsupervised updating of land-cover maps

Bruzzone

Cossu

2002

IEEE Trans. Geosci. Remote Sensing

112

View full text Add to dashboard Cite

show abstract

“…Multiple expert fusion aims to make use of many different designs to improve the classification performance. Over the last few years a myriad of methods for fusing the outputs of multiple classifiers have been proposed [34,35,36,37,38,39,40,41,42,43,44]. The methods range from simple Bayesian estimations methods, through trainable multistage strategies where the outputs of component classifiers are considered as features and the fusion is performed by another classifier designed using independent data, to data dependent methods where each classifier has a domain of superior competence and their opinion is called on only when the observation falls into this domain.…”

Section: Introductionmentioning

confidence: 99%

A Framework for Classifier Fusion: Is It Still Needed?

Kittler

2000

Advances in Pattern Recognition

View full text Add to dashboard Cite

Abstract. We consider the problem and issues of classifier fusion and discuss how they should be reflected in the fusion system architecture. We adopt the Bayesian viewpoint and show how this leads to classifier output moderation to compensate for sampling problems. We then discuss how the moderated outputs should be combined to reflect the prior distribution of models underlying the classifier designs. We then elaborate how the final stage of fusion should combine the complementary measurement information that might be available to different experts. This process is embodied in an overall architecture which shows why the fusion of raw expert outputs is a nonlinear function of the expert outputs and how this function can be realised as a sequence of relatively simple processes.

show abstract

“…An example of score level fusion methods is the Bayesian Fusion. In [58,59], the authors introduced two statistical frameworks for score level fusion of classifiers based on the Bayesian theorem. The first framework proposed a fusion method for classifiers that use distinct representation in the input layer.…”

Section: Score Level Fusionmentioning

confidence: 99%

Temporal contextual descriptors and applications to emotion analysis.

Balti¹

View full text Add to dashboard Cite

The current trends in technology suggest that the next generation of services and devices allows smarter customization and automatic context recognition. Computers learn the behavior of the users and can offer them customized services depending on the context, location, and preferences.One of the most important challenges in human-machine interaction is the proper understanding of human emotions by machines and automated systems. In the recent years, the progress made in machine learning and pattern recognition led to the development of algorithms that are able to learn the detection and identification of human emotions from experience. These algorithms use different modalities such as image, speech, and physiological signals to analyze and learn human emotions. In many settings, the vocal information might be more available than other modalities due to widespread of voice sensors in phones, cars, and computer systems in general.In emotion analysis from speech, an audio utterance is represented by an ordered (in time) sequence of features or a multivariate time series. Typically, the sequence is further mapped into a global descriptor representative of the entire utterance/sequence. This descriptor is used for classification and analysis. In classic approaches, statistics iii are computed over the entire sequence and used as a global descriptor. This often results in the loss of temporal ordering from the original sequence. Emotion is a succession of acoustic events. By discarding the temporal ordering of these events in the mapping, the classic approaches cannot detect acoustic patterns that lead to a certain emotion.In this dissertation, we propose a novel feature mapping framework. The proposed framework maps temporally ordered sequence of acoustic features into data-driven global descriptors that integrate the temporal information from the original sequence.The framework contains three mapping algorithms. These algorithms integrate the temporal information implicitly and explicitly in the descriptor's representation.In the first algorithm, the Temporal Averaging Algorithm, we average the data temporally using leaky integrators to produce a global descriptor that implicitly integrates the temporal information from the original sequence.In order to integrate the discrimination between classes in the mapping, we propose the Temporal Response Averaging Algorithm which combines the temporal averaging step of the previous algorithm and unsupervised learning to produce data driven temporal contextual descriptors.In the third algorithm, we use the topology preserving property of the Self-Organizing Maps and the continuous nature of speech to map a temporal sequence into an ordered trajectory representing the behavior over time of the input utterance on a 2-D map of emotions. The temporal information is integrated explicitly in the descriptor which makes it easier to monitor emotions in long speeches.The proposed mapping framework maps speech data of different length to the same equivalent representation which allevi...

show abstract

Strategies for combining classifiers employing shared and distinct pattern representations

Cited by 34 publications

References 4 publications

A multiple-cascade-classifier system for a robust and partially unsupervised updating of land-cover maps

A multiple-cascade-classifier system for a robust and partially unsupervised updating of land-cover maps

A Framework for Classifier Fusion: Is It Still Needed?

Temporal contextual descriptors and applications to emotion analysis.

Contact Info

Product

Resources

About