Influence of erroneous learning samples on adaptation in on-line handwriting recognition

Pattern Recognition

Volumn 35, Issue 4, 2002, Pages 915-925

  Vuori, Vuokko ^a   Laaksonen, Jorma ^a   Kangas, Jari ^b  

^a AALTO UNIVERSITY   (Finland)

^b NOKIA RESEARCH CENTER   (United States)

Author keywords

Adaptation; Dynamic time warping (DTW); Erroneous learning samples; Intelligent user interface; Isolated character recognition; K nearest neighbor rule; Learning vector quantization (LVQ); On line recognition; Unconstrained writing style

Indexed keywords

ALGORITHMS; COMPUTER SIMULATION; ONLINE SYSTEMS; VECTORS;

ON-LINE HANDWRITING RECOGNITION SYSTEMS;

CHARACTER RECOGNITION;

PATTERN RECOGNITION;

EID: 0036532795     PISSN: 00313203     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0031-3203(01)00083-8     Document Type: Article

References (11)

1. Integration of lexical and syntactical knowledge in a handwriting-recognition system
2. Writer independent on-line handwriting recognition using an HMM approach
3. Alphanumeric entry on pen-based computers
4. Adaptation of prototype sets in on-line recognition of isolated handwritten latin characters
5. Comparison of adaptive strategies for on-line character recognition
6. UNIPEN project of on-line data exchange and recognizer benchmark
7. An important contribution to nonparametric discriminant analysis and density estimation-Commentary on Fix and Hodges (1951)