Word boundary detection using pitch variations
نویسندگان
چکیده
This paper proposes a method for detecting word boundaries. This method is based on the behaviour of the pitch frequency across the sentences. The pitch frequency (F0) is found to rise in a word and fall to the next word. The presence of this fall is proposed as a means of detecting word boundaries. Four major Indian languages are used and the results show that nearly 85% of the word boundaries were correctly detected. The same method used for German language shows that nearly 65% of the word boundaries were correctly detected. The implications of these result in the development of a continuous speech recognition system are discussed.
منابع مشابه
Word segmentation in Persian continuous speech using F0 contour
Word segmentation in continuous speech is a complex cognitive process. Previous research on spoken word segmentation has revealed that in fixed-stress languages, listeners use acoustic cues to stress to de-segment speech into words. It has been further assumed that stress in non-final or non-initial position hinders the demarcative function of this prosodic factor. In Persian, stress is retract...
متن کاملWord Boundary Detection Using Pitch Variations
This paper proposes a method for detecting word boundaries. This method is based on the behaviour of the pitch frequency across the sentences. The pitch frequency (F0) is found to rise in a word and fall to the next word. The presence of this fall is proposed as a means of detecting word boundaries. Four major Indian languages are used and the results show that nearly 85% of the word boundaries...
متن کاملDirector Structures in a Chiral Nematic Slab: Threshold Field and Pitch Variations
Abstract The liquid crystal director distribution is determined for a confined chiral nematic slab. The molecular director distribution of the field-controlled chiral nematic slab is directly calculated. The director profiles for the tilt and the twist angles, under different applied fields, are calculated in the slab with weak boundary conditions. Then, the dependence of the threshold field on...
متن کاملCS281B Spring 2004 – Project Report Bagged decision trees vs. Bayesian neural network for sentence boundary detection
The task is to detect sentence boundaries in conversational speech. This is done at the word level, using mainly prosodic features (based on patterns of pitch, duration, and stress). There are a few non-prosodic features such as an estimate of the speaker gender and whether there is a speaker change at that word. (Speaker change does not always signal a sentence boundary, since a second speaker...
متن کاملThe role of post-lexical tonal contou
The current study examines the role of non-lexical tonal patterns of a prosodic phrase in word segmentation. Seventy-two native Seoul Korean speakers participated in the word spotting experiments. The results showed that listeners spotted more words in phrase initial than in phrase medial position, and that their error rates were significantly lower in rising tonal patterns than in non-rising p...
متن کامل