Search

Prosody development provides insight into the development of a highly specialized system that is unique to humans. Prosody is mainly encoded by regulating the fundamental frequency, intensity and duration of various tokens during discourse construction. Interplay between the fundamental frequency and the intensity plays a crucial role in formulating prosody and adult speakers modulate these parameters to produce rhythmic sentences. However, it is not known, how this relationship between these two parameters evolve with age in the children. The present study tries to fill this gap partially by investigating the statistical nature of the correlation between fundamental frequency and intensity in the speech of 42 children and 25 adults between the ages of 4–25 years. Statistically significant differences were found in the distributions of correlations between children and adults. The difference can be attributed to evolving skills of children in encoding prosodic cues through modulation of intensity by alterations in flow of air from the lungs and controlling the vibrations of vocal folds.

Spectro-temporal features of speech are the basis of phonological comprehension and production in the brain. Thus, these features provide a relevant framework to study speech and language development in children. In this paper, we present a novel framework to study the statistics of spectro-temporal features of speech that are encoded at different timescales. These timescales correspond to different linguistic units such as prosodic or syllabic components. The framework is tested on a speech corpus consisting of 169 speech samples. The paper demonstrates usage of the proposed framework in finding milestones of speech development in children. The results indicate the presence of more number of spectro-temporal features encoded at short timescales in adults as compared to children. However, no significant difference is observed in the spectro-temporal features encoded at long timescales between these groups. The proposed framework is also used in studying the speech impairments of children and adults with mild to moderate intellectual disabilities. The results reveal the absence of some spectro-temporal features encoded at both the timescales and their absence is more prominent in shorter timescales. The suggested framework can be used for studying speech development and impairment in different disorders.

This study presents a geometric feature based automatic facial expression recognition system. The proposed system utilises the facial landmark points to determine the relative distances between the facial features in order to capture deformities caused by the movement of facial muscles due to different expressions. Three feature sets are generated by using landmark coordinates, relative distances between the facial points and a combination of both. Discriminating power of each feature set is determined by training different classification models for classifying an image into six basic emotions or neutral state. The proposed system is validated on two publically available facial expression databases. Experimental results show good accuracy of 95.5% for MUG database on the combined features by using ensemble neural network.

Indian children represent a unique population of bilinguals who learn two or more languages of different isochrony. Most of their native languages are syllable-timed, whereas in school the language of learning is English, which is a stress-timed language. Exposure to multiple languages with different types of rhythmic patterns as stimulus during the development stage differentiates these children from monolinguals acquiring rhythms of a single language. Thus, to partially fill this gap, a comparative study in 115 children (aged four to eight years) and 18 adults was carried out. We investigated the frequencies around the pitch and first formant which encodes prosodic information in speech. It was found that the harmonic content in speech increases even after four years (where children exhibit or possess intelligent speech) and the spectra of adults were more distributed and had more peaks at higher frequencies as compared to the power spectra of children, where energy in higher frequency was less. This study has implications in finding speech abnormalities in Indian children with speech disorders.