Speech recognition systems for languages with a rich in ectional morphology (like German) suffier from the limitations of a word-based full-form lexicon. Although the morphological and acoustical knowledge about words is coded implicitly within the lexicon entries (which are usually closely related to the orthography of the language at hand) this knowledge is usually not explicitly available for other tasks (e.g. detecting OOV words). This paper presents an HMM-based `word' recognizer that uses morphemes on the string level for recognizing spontaneous German conversational speech ( Verbmobil corpus). The system has no explicit word knowledge but uses a morpheme{bigram to capture the German word and sentence structure to some extent. The morpheme recognizer is tightly coupled with a prosodic classifier in order to compensate for some of the additional ambiguity introduced by using morphemes instead of words. Although the recognizer's morpheme accuracy of 85:3% is comparable to that of our word-based decoder (word accuracy 86%) until now the benefit of introducing the prosodic classifier is not yet clear.
Cite as: Strom, V., Heine, H. (1999) Utilizing prosody for unconstrained morpheme recognition. Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999), 307-310, doi: 10.21437/Eurospeech.1999-80
@inproceedings{strom99_eurospeech,
author={Volker Strom and Henrik Heine},
title={{Utilizing prosody for unconstrained morpheme recognition}},
year=1999,
booktitle={Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999)},
pages={307--310},
doi={10.21437/Eurospeech.1999-80}
}