5th International Conference on Spoken Language Processing
There are many Spontaneous Dialogue Recognition based applications like home-banking ones where long numbers recognition facilities are crucial to complete a request from the user. Rejection and Utterance Verification (UV) are difficult problems in these applications. In this contribution we improve our previously proposed UV procedure in order to increase the correction of recognition errors, to solve grammatical ambiguities from the user, and to make more efficient the rejection of misrecognized or out-of-vocabulary (OOV) utterances. In spite of the verification performance, the proposed algorithm complies with the real-time constrains which are mandatory in real applications. We evaluate our method and present recognition results from the long natural number recognition task of a real Data Driven application through the telephone line on a multilingual environment. Experimental results show that the proposed method obtains a significant reduction in terms of recognition errors and achieves an extraordinary low false acceptance rate in all cases for different languages.
Bibliographic reference. Caminero, Javier / López, Eduardo / Hernández, Luis A. (1998): "Two-pass utterance verification algorithm for long natural numbers recognition", In ICSLP-1998, paper 0440.