Recurrent Out-of-Vocabulary Word Detection Using Distribution of Features
Taichi Asami, Ryo Masumura, Yushi Aono, Koichi Shinoda · 2016
The repeated use of out-of-vocabulary (OOV) words in a spo- ken document seriously degrades a speech recognizer’s perfor- mance. This paper provides a novel method for accurately de- tecting such recurrent OOV words. Standard OOV word de- tection methods classify each word segment into in-vocabulary (IV) or OOV. This word-by-word classification tends to be af- fected by sudden vocal irregularities in spontaneous speech, triggering false alarms. To avoid this sensitivity to the irreg- ularities, our proposal focuses on consistency of the repeated occurrence of OOV words. The proposed method preliminar- ily detects recurrent segments, segments that contain the same word, in a spoken document by open vocabulary spoken term discovery using a phoneme recognizer. If the recurrent seg- ments are OOV words, features for OOV detection in those segments should exhibit consistency. We capture this consis- tency by using the mean and variance (distribution) of features (DOF) derived from the recurrent segments, and use the DOF for IV/OOV classification. Experiments illustrate that the pro- posed method’s use of the DOF significantly improves its per- formance in recurrent OOV word detection. Index Terms: speech recognition, OOV word detection, recur- rent OOV words, distribution of features