J. Laaksonen, V. Vuori, E. Oja, and J. Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters. In Seong-Whan Lee, editor, Advances in Handwriting Recognition, pages 489--497. World Scientific Publishing, 1999. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Influence of Erroneous Learning Samples on Adaptation in .. - Vuori, Laaksonen, Kangas (2001) (1 citation) Self-citation (Laaksonen Vuori Kangas) (Correct)
....comparing the input characters and prototypes, and a decision rule according to which the classifications are carried out. The prototype set was formed by clustering character samples written by 22 subjects and selecting the center most item of each cluster as a prototype. The clustering algorithm [4] was semiautomatic as the number of clusters per each class had to be predefined. The number of clusters, or prototypes, was set to seven for all the character classes as it was the maximum number of different writing styles found for a single class in the manual examination of the data. The k ....
Jorma Laaksonen, Vuokko Vuori, Erkki Oja, and Jari Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters. In Seong-Whan Lee, editor, Advances in Handwriting Recognition, pages 489--497. World Scientific Publishing, 1999.
Dynamically Expanding Context as Committee Adaptation.. - Laaksonen, Aksela, Oja (1999) Self-citation (Laaksonen Oja Kangas) (Correct)
....characters were used to form the user independent member classifiers of the committee. There were four classifiers, each based on point wise elastic matching [5] between the input character and a set of 273 stored prototypes. The prototypes were selected with a semiautomatic clustering algorithm [4]. The difference between the four classifiers was in the normalization of the characters. In two of them, the characters were centered by their bounding boxes and in the other two, by their mass centers. On the other hand, in two classifiers the size of the characters was normalized and in the ....
....have described a handwriting recognition system in which a set of static classifiers is organized in a committee with adaptive decision rules. The results of the performed experiments show that this approach produces recognition results comparable with our earlier studies with adaptive classifiers [3, 4, 7]. The principle of DEC, the Dynamically Expanding Context, was thus proven to be suited for the implementation of adaptive symbol transformation rules also in the case of handwriting recognition. 0 100 200 300 400 500 600 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 recognition ....
J. Laaksonen, V. Vuori, E. Oja, and J. Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten Latin characters. In S.-W. Lee, editor, Advances in Handwriting Recognition, pages 489--497. World Scientific Publishing, 1999.
Controlling On-Line Adaptation of a Prototype-Based.. - Vuori, Laaksonen, Oja (2000) Self-citation (Laaksonen Vuori Oja Kangas) (Correct)
....The classifier used in the experiments is based on the k nearest neighbor (k NN) rule: an input character is matched against all stored prototypes and the k most similar ones vote for its classification. The prototype set covers several writing styles and it is formed with a clustering algorithm [2] from a database including characters written by several subjects. The number of prototypes per class is always seven, which was the maximum number of different writing styles found for a single character in the manual examination of the data. The dissimilarity measure [6] used both for the ....
J. Laaksonen, V. Vuori, E. Oja, and J. Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters. In S.-W. Lee, editor, Advances in Handwriting Recognition, pages 489--497. World Scientific Publishing, 1999.
Adaptive Local Subspace Classifier in on-line recognition .. - Laaksonen, Aksela, Oja (1999) Self-citation (Laaksonen Oja Kangas) (Correct)
....we have used is the Local Subspace Classifier (LSC) of which we have developed an adaptive version. Introduction We have developed an on line recognition system for handwritten characters which is aimed at being used in personal digital assistants (PDAs) and other portable handheld devices [3]. The system applies continuous adaptation to the user s writing style. The adaptation takes place simultaneously with the normal operation of the system and, therefore, there is no need for a separate training period of the device. Character recognition in our system is currently performed by ....
J. Laaksonen, V. Vuori, E. Oja, and J. Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten Latin characters. In S.-W. Lee, editor, Advances in Handwriting Recognition. World Scientific Publishing, 1999.
Adaptive Character Recognizer For A Hand-Held Device.. - Vuori, Aksela.. (2000) (2 citations) Self-citation (Laaksonen Vuori Oja Kangas) (Correct)
No context found.
Jorma Laaksonen, Vuokko Vuori, Erkki Oja, and Jari Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters. In Seong-Whan Lee, editor, Advances in Handwriting Recognition, pages 489497. World Scientic Publishing, 1999.
Experiments with Adaptation Methods in On-line.. - Laaksonen, Vuori.. (2000) Self-citation (Laaksonen Vuori Oja Kangas) (Correct)
....In these runs, the number of prototypes was the same for every class, namely seven. The number of clusters per stroke number variation was selected so that it roughly corresponds to the respective share of all the writing styles of that character. The algorithm is explained in detail in [11]. The second clustering algorithm utilizes the opposite approach. At the beginning of the algorithm, there are as many clusters as there are character samples. Next, those two cluster whose middle items are the most similar are merged and the middle item 11 of the new cluster is found. Then ....
Jorma Laaksonen, Vuokko Vuori, Erkki Oja, and Jari Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters. In Seong-Whan Lee, editor, Advances in Handwriting Recognition, pages 489--497. World Scientific Publishing, 1999.
Dynamic Time Warping: An intuitive way of handwriting recognition? - Niels (2004) (Correct)
No context found.
J. Laaksonen, V. Vuori, E. Oja, and J. Kangas. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters. In Seong-Whan Lee, editor, Advances in Handwriting Recognition, pages 489--497. World Scientific Publishing, 1999.
Online Recognition of Chinese Characters: The State-of-the-Art - Liu, Jaeger, Nakagawa (Correct)
No context found.
J. Laaksonen, V. Vuori, and E. Oja, "Adaptation of Prototype Sets in On-Line Recognition of Isolated Handwritten Latin Characters, " Advances in Handwriting Recognition, S.-W. Lee ed., pp. 489497, World Scientific, 1999.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC