Appeal No. 1998-1386 Page 20 Application No. 08/269,156 Fujisaki does teach considering both character shape and context information. Regarding the former, the reference mentions, “a Frame Shape Recognizer, associates conditional probabilities to each frame ....” Col. 13, ll. 67-68. Regarding the latter, the reference states, “The language model block 26 provides language model probabilities which may be used to determine what characters are most likely to occur in a given context, or what words are most likely to occur in a given context.” Col. 9, ll. 23-27. The examiner does not show, however, that the Frame Shape Recognizer and the language model block operate substantially simultaneously, let alone simultaneously. To the contrary, Fujisaki suggests that the components operate sequentially. Specifically, the reference characterizes the Frame Shape Recognizer as one of the “pre-processing components ....” Col. 13, ll. 63-66. Such a characterization implies that the Frame Shape Recognizer operates before the processing of the language model block.Page: Previous 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 NextLast modified: November 3, 2007