An HMM Based Two-Pass Approach for Off-Line Cursive Handwriting Recognition
ICMI '00 Proceedings of the Third International Conference on Advances in Multimodal Interfaces
Motion intention recognition in robot assisted applications
Robotics and Autonomous Systems
ICANN'07 Proceedings of the 17th international conference on Artificial neural networks
Data-embedding pen: augmenting ink strokes with meta-information
DAS '10 Proceedings of the 9th IAPR International Workshop on Document Analysis Systems
Hi-index | 0.00 |
This paper deals with the performance evaluation of a novel hybrid approach to large vocabulary cursive handwriting recognition and contains the following innovations. 1) It presents the investigation of a new hybrid approach to handwriting recognition, consisting of Hidden Markov Models (HMMs) and neural networks trained with a special information theory-based training criterion. This approach has only been recently introduced successfully to on-line handwriting recognition and is now investigated for the first time for off-line recognition. 2) The hybrid approach is extensively compared to traditional HMM-modeling techniques and the superior performance of the new hybrid approach is demonstrated. 3) The data for the comparison has been obtained from a database containing on-line handwritten data which has been converted to off-line data. Therefore, a multiple evaluation has been carried out, incorporating the comparison of different modeling techniques and the additional comparison of each technique for on-line and off-line recognition, using a unique database. The results confirm the fact that on-line recognition leads to better recognition results due to the dynamic information of the data, but also show that it is possible to obtain recognition rates for off-line recognition that are close to the results obtained for on-line recognition. Furthermore, it can be shown that for both, on-line and off-line recognition, the new hybrid approach clearly outperforms the competing traditional HMM techniques. It is also demonstrated that the new hybrid approach yields superior results for the off-line recognition of machine-printed multifont characters.