Fast Online Incremental Learning with Few Examples For Online Handwritten Character Recognition
Abdullah Almaksour, Harold Mouchère, Éric Anquetil · 2008
An incremental learning strategy for handwritten character recognition is proposed in this paper. The strat-egy is online and fast, in the sense that any new char-acter class can be instantly learned by the system. The proposed strategy aims at overcoming the problem of lack of training data when introducing a new character class. Synthetic handwritten characters generation is used for this purpose. Our approach uses a Fuzzy Inference Sys-tem (FIS) as a classifier. Results have shown that a good recognition rate (about 90%) can be achieved using only 3 training examples. And such rate rapidly improves reach-ing 96 % for 10 examples, and 97 % for 30 ones.