Publication Details
Out-of-Vocabulary Word Detection and Beyond
Hannemann Mirko, Dipl.-Ing. (DCGM FIT BUT)
Burget Lukáš, doc. Ing., Ph.D. (DCGM FIT BUT)
out-of-vocabulary word, detection, neural network, speech recognition
In this work, we investigate into two approaches for OOV word detection. We compare both systems in a fusion experiment, and describe how to actually make use of the detected incongruence.
In this work, we summarize our experiences in detection of unexpected words in automatic speech recognition (ASR). Two ap- proaches based upon a paradigm of incongruence detection between generic and specific recognition systems are introduced. By arguing, that detec- tion of incongruence is a necessity, but does not suffice when having in mind possible follow-up actions, we motivate the preference of one ap- proach over the other. Nevertheless, we show, that a fusion outperforms both single systems. Finally, we propose possible actions after the de- tection of unexpected words, and conclude with general remarks about what we found to be important when dealing with unexpected words.
@INBOOK{FITPUB9944, author = "Stefan Kombrink and Mirko Hannemann and Luk\'{a}\v{s} Burget", title = "Out-of-Vocabulary Word Detection and Beyond", pages = "57--65", booktitle = "Detection and Identification of Rare Audiovisual Cues", series = "Studies in Computational Intelligence, 384", year = 2012, location = "Springer-Verlag Berlin Heidelberg, DE", publisher = "Springer Verlag", ISBN = "978-3-642-24033-1", doi = "10.1007/978-3-642-24034-8\_4", language = "english", url = "https://www.fit.vut.cz/research/publication/9944" }