Browsing by Author "Uhlmann S."

Now showing 1 - 4 of 4

Citation - Scopus: 5
Collective Network of Evolutionary Binary Classifiers for Content-Based Image Retrieval
(2011) Kiranyaz S.; Uhlmann S.; Pulkkinen J.; Gabbouj, Moncef; İnce, Türker
The content-based image retrieval (CBIR) has been an active research field for which several feature extraction, classification and retrieval techniques have been proposed up to date. However, when the database size grows larger, it is a common fact that the overall retrieval performance significantly deteriorates. In this paper, we propose collective network of (evolutionary) binary classifiers (CNBC) framework to achieve a high retrieval performance even though the training (ground truth) data may not be entirely present from the beginning and thus the system can only be trained incrementally. The CNBC framework basically adopts a "Divide and Conquer" type approach by allocating several networks of binary classifiers (NBCs) to discriminate each class and performs evolutionary search to find the optimal binary classifier (BC) in each NBC. In such an evolution session, the CNBC body can further dynamically adapt itself with each new incoming class/feature set without a full-scale re-training or re-configuration. Both visual and numerical performance evaluations of the proposed framework over benchmark image databases demonstrate its scalability; and a significant performance improvement is achieved over traditional retrieval techniques. © 2011 IEEE.
Incremental Evolution of Collective Network of Binary Classifier for Content-Based Image Classification and Retrieval
(2011) Kiranyaz S.; Uhlmann S.; Pulkkinen J.; İnce, Türker; Gabbouj M.
In this paper, we propose an incremental evolution scheme within collective network of (evolutionary) binary classifiers (CNBC) framework to address the problem of incremental learning and to achieve a high retrieval performance for content-based image retrieval (CBIR). The proposed CNBC framework can still function even though the training (ground truth) data may not be entirely present from the beginning and thus the system can only be evolved incrementally. The CNBC framework basically adopts a "Divide and Conquer" type approach by allocating several networks of binary classifiers (NBCs) to discriminate each class and performs evolutionary search to find the optimal binary classifier (BC) in each NBC. This design further allows such scalability that the CNBC can dynamically adapt its internal topology to new features and classes with minimal effort. Both visual and numerical performance evaluations of the proposed framework over benchmark image databases demonstrate its efficiency and accuracy for scalable CBIR and classification. © 2011 IEEE.
Citation - Scopus: 7
Polarimetric Sar Images Classification Using Collective Network of Binary Classifiers
(2011) Uhlmann S.; Kiranyaz S.; Gabbouj, Moncef; İnce, Türker
In this paper, we propose the application of collective network of (evolutionary) binary classifiers (CNBC) to address the problems of feature/class scalability and classifier evolution, to achieve a high classification performance over full polarimetric SAR images even though the training (ground truth) data may not be entirely accurate. The CNBC basically adopts a "Divide and Conquer" type approach by allocating an individual network of binary classifiers (NBCs) to discriminate each SAR image class and performing evolutionary search to find the optimal binary classifier (BC) in each NBC. Such design further allows dynamic class and SAR image feature scalability in such a way that the CNBC can gradually adapt itself to new features and classes with minimal effort. Experiments demonstrate the classification accuracy and efficiency of the proposed system over the fully polarimetric AIRSAR San Francisco Bay data set. © 2011 IEEE.
Citation - WoS: 2
Citation - Scopus: 2
Sar Imagery Classification in Extended Feature Space by Collective Network of Binary Classifiers
(2011) Uhlmann S.; Kiranyaz S.; İnce, Türker; Gabbouj, Moncef
Polarimetric SAR image classification has been an active research field where several features and classifiers have been proposed in the past. Using numerous features can be a desirable option so as to achieve a better discrimination over certain classes, yet key questions such as how to avoid "Curse of Dimensionality" and how to combine them in the most effective way still remains unanswered. In this paper, we investigate SAR image classification using a large set of features, where the focus is particularly drawn on the extension of image processing features e.g. texture, edge and color. We propose a dedicated application of the Collective Network of (evolutionary) Binary Classifiers (CNBC) framework to address these problems with the aim of achieving high feature scalability. We furthermore tested several SAR and image processing feature constellations over three well-known SAR image classifiers and make comparative evaluations with CNBC. Experimental results over the full polarimetric AIRSAR San Francisco Bay and Flevoland images show that additional image processing features are able to improve SAR image classification accuracy and moreover, the CNBC proves useful and can scale well especially whenever high number of features and classes are encountered. © 2011 EURASIP.