On the Use of Different Speech Representations for Speaker Modeling

Chen, Ke

doi:10.1109/tsmcc.2005.848166

Cited by 20 publications

(11 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…3, it is also observed that the semi-supervised learning system of a strong classifier merely yields an improvement only when the ICA representation is used but fails to work on other two representations. By comparison, we conclude that the simultaneous use of different representations results in robust learning and better generalization during semi-supervised learning, which is completely consistent with our previous argument in unsupervised [13] and supervised learning [14], [15].…”

Section: Facial Expression Recognitionsupporting

confidence: 90%

“…One is that the nature of the task makes different redundant representations 1 available, and the other is no different representations available. For the former case, our earlier studies show the usefulness of combining multiple classifiers trained on different representations [13]- [15]. For the latter case, we can use the bootstrap re-sampling techniques [16] to create different data sets and then train an ensemble of classifiers on them.…”

Section: B Classifier Ensemble Generationmentioning

confidence: 99%

See 1 more Smart Citation

Ensemble Learning with Active Data Selection for Semi-Supervised Pattern Classification

Wang

Chen

2007

2007 International Joint Conference on Neural Networks

View full text Add to dashboard Cite

Unliketraditional pattern classification, semi-supervised learning provides a novel technique to make use of both labeled and unlabeled data for improving the performance of classification. In general, there are two critical issues for semi-supervised learning of discriminative classifiers; i.e., how to create an initial classifier of a good generalization capability with the limited labeled data and the how to make an effective use of unlabeled data without degradation of the established classifier. To tackle two aforementioned problems, we propose an ensemble learning approach based on a recent active data selection strategy [1], where ensemble learning would yield good generalization and active data selection tends to choose the unlabeled data more likely resulting in an improvement during semi-supervised learning. By using an ensemble of K-NN classifiers, we demonstrate the effectiveness of our approach on a synthetic data classification and a facial expression recognition tasks.

show abstract

Section: Facial Expression Recognitionsupporting

confidence: 90%

Section: B Classifier Ensemble Generationmentioning

confidence: 99%

Ensemble Learning with Active Data Selection for Semi-Supervised Pattern Classification

Wang

Chen

2007

2007 International Joint Conference on Neural Networks

View full text Add to dashboard Cite

show abstract

“…İkinci bir yöntem, farklı özniteliklerin farklı sınıflayıcılarla beraber kullanılarak işlenmesi yöntemidir [44,45]. Yaygın olarak kullanılmakta olan bu ikinci yolda önemli olan, en uygun öznitelik-sınıflayıcı eşleşmelerini yapabilmektir.…”

Section: Gi̇ri̇ş (Introduction)unclassified

Batin Bölgesi̇ Organlarinin MR Görüntüleri̇nden Çok Aşamali Hi̇yerarşi̇k Siniflama İle Bölütlenmesi̇

Selvi¹,

Selver²,

Kavur³

et al. 2015

Gazi Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi

View full text Add to dashboard Cite

ÖZETTıbbi görüntüleme ile anatomi hakkında ayrıntılı bilgi elde edinilebildiğinden, tanı amaçlı görüntüleme günümüzde birçok açıdan önem kazanmıştır. Görüntüleme cihazları tarafından sunulan verilerin fazlalığı ve çeşitliliği nedeniyle, tüm veri yerine görüntülerde sadece ilgilenilen dokunun belirlenerek ayrılması (Bölütlenmesi) sağlanabilir. Elcil yöntemler ile bölütleme yorucu, zaman alıcı ve deneyim gerektiren bir işlem olduğundan, otomatik yordamlara gereksinim duyulmaktadır. Geliştirilen yordamların klinik koşullarında kullanılabilmesi içinse yüksek başarıma sahip sonuçlar üretmeleri gerekmektedir. Manyetik Rezonans (MR) görüntülerinden batın bölgesindeki organlarının bölütlenmesi pek çok zorluk içeren bir uygulama alanıdır ve bu konudaki çalışmalar sınırlı sayıdadır. Batın bölgesinde yer alan, karaciğer, böbrekler, dalak, pankreas, safra kesesi gibi organların MR görüntüleri kullanılarak ileri seviye tıbbi analizi ve üç boyutlu incelenmesi pek çok tıbbi prosedür için mecburi olduğundan, bu çalışmada, ilgili organların bölütlenmesinde yukarıda belirtilen zorluklara karşı gürbüz, bölütlenecek organın özellikleri ve organların birbirleriyle olan ilişkilerini (konum vb.) göz önüne alan bir sistem geliştirilmiştir. Geliştirilen sistem farklı MR sekansları ile elde edilen görüntülere uygulanarak elde edilen sonuçlar tartışılmıştır.Anahtar Kelimeler: Bölütleme, MR, hiyerarşik sınıflama, batın SEGMENTATION OF ABDOMINAL ORGANS FROM MR IMAGES USING MULTI-LEVEL HIERARCHICAL CLASSIFICATION ABSTRACTMedical imaging modalities can provide very detailed and informative mappings of the anatomy of a subject. These detailed and informative mappings can be processed to extract the information of interest instead of dealing with whole data (segmentation). Since manual segmentation on each slice is time consuming, tedious and operator dependent, automatic tools and techniques are needed. Segmentation of abdominal organs is a very challenging field of application due to overlapping intensity ranges of the organs, variations in human anatomy and pathology and the number of studies is very limited for Magnetic Resonance (MR), which is a relatively newer and rapidly developing imaging modality.Since it is obligatory to analyze and visualize MR images of abdominal organs (i.e. liver, right/left kidneys, spleen, pancreas, gall bladder) for several medical procedures, the main goal of this paper is to design and develop a segmentation system (method+software), which is robust to the challenges mentioned above, adaptive to the properties of the abdominal organs as well as to the interrelationships of these organs.

show abstract

“…Unlike the previous method [20], our approach is motivated by our previous success in the use of different representations to construct an ensemble model for dealing with difficult supervised [22]- [24] and semisupervised learning tasks [25], where the use of different representations better exploits the information conveyed in the raw data and therefore leads to the better performance. For each individual representation, we first employ an RPCL network for clustering analysis of automatic model selection, and the nature of an RPCL network often leads to quick clustering analysis.…”

Section: Introductionmentioning

confidence: 99%

Time Series Clustering Via RPCL Network Ensemble With Different Representations

Yang

Chen

2011

IEEE Trans. Syst., Man, Cybern. C

View full text Add to dashboard Cite

Abstract-Time series clustering provides underpinning techniques for discovering the intrinsic structure and condensing/summarizing information conveyed in time series, which is demanded in various fields ranging from bioinformatics to video content understanding. In this paper, we present an unsupervised ensemble learning approach to time series clustering by combining rival-penalized competitive learning (RPCL) networks with different representations of time series. In our approach, the RPCL network ensemble is employed for clustering analyses based on different representations of time series whenever available, and an optimal selection function is applied to find out a final consensus partition from multiple partition candidates yielded by applying various consensus functions for the combination of competitive learning results. As a result, our approach first exploits its capability of the RPCL rule in clustering analysis of automatic model selection on individual representations and subsequently applies ensemble learning for the synergy of reconciling diverse partitions resulted from the use of different representations and augmenting RPCL networks in automatic model selection and overcoming its inherent limitation. Our approach has been evaluated on 16 benchmark time series data mining tasks with comparison to state-of-the-art time series clustering techniques. Simulation results demonstrate that our approach yields favorite results in clustering analysis of automatic model selection.

show abstract

On the Use of Different Speech Representations for Speaker Modeling

Cited by 20 publications

References 44 publications

Ensemble Learning with Active Data Selection for Semi-Supervised Pattern Classification

Ensemble Learning with Active Data Selection for Semi-Supervised Pattern Classification

Batin Bölgesi̇ Organlarinin MR Görüntüleri̇nden Çok Aşamali Hi̇yerarşi̇k Siniflama İle Bölütlenmesi̇

Time Series Clustering Via RPCL Network Ensemble With Different Representations

Contact Info

Product

Resources

About