Combining DEKF algorithm and trace rule for fast on-line invariance extraction and recognition
Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review
Author(s)
Related Research Unit(s)
Detail(s)
Original language | English |
---|---|
Pages (from-to) | 493-501 |
Journal / Publication | Pattern Recognition Letters |
Volume | 21 |
Issue number | 6-7 |
Publication status | Published - Jun 2000 |
Link(s)
Abstract
A self-organizing network is used to perform invariance extraction and recognition of handwritten digits. To extract the invariance effectively, we propose to combine the trace learning rule and the on-line dual extended Kalman filter (DEKF) algorithm. Furthermore, a new activation function is suggested to replace the traditional sigmoid activation function so as to reduce the sensitivity of the extracted features to samples with large variance. Computer simulations show that both the learning speed and the recognition rate are improved using a compact network. © 2000 Elsevier Science B.V.
Research Area(s)
- EKF Algorithm, Neural networks, Pattern recognition, Pruning, Trace Rule
Citation Format(s)
Combining DEKF algorithm and trace rule for fast on-line invariance extraction and recognition. / Chang, Sheng-Jiang; Wong, Kwok-Wo; Leung, Chi-Sing.
In: Pattern Recognition Letters, Vol. 21, No. 6-7, 06.2000, p. 493-501.
In: Pattern Recognition Letters, Vol. 21, No. 6-7, 06.2000, p. 493-501.
Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review