Estimating velum height from acoustics during continuous speech
نویسنده
چکیده
This paper reports on present work, in which a recurrent neural network is trained to estimate ‘velum height’ during continuous speech. Parallel acoustic-articulatory data comprising more than 400 read TIMIT sentences is obtained using electromagnetic articulography (EMA). This data is processed and used as training data for a range of neural network sizes. The network demonstrating the highest accuracy is identified. This performance is then evaluated in detail by analysing the network’s output for each phonetic segment contained in 50 hand-labelled utterances set aside for testing purposes.
منابع مشابه
Estimation of articulatory gesture patterns from speech acoustics
We investigated dynamic programming (DP) and statemodel (SM) approaches for estimating gestural scores from speech acoustics. We performed a word-identification task using the gestural pattern vector sequences estimated by each approach. For a set of 75 randomly chosen words, we obtained the best word-identification accuracy (66.67%) using the DP approach. This result implies that considerable ...
متن کاملMotor equivalent strategies in the production of /u/ in perturbed speech
Several articulatory strategies are available during the production of /u/, all resulting in a similar acoustic output. /u/ has two main constrictions, at the velum and at the lips. A perturbation of either constriction can be compensated at the other one, e.g wider constriction at the velum by more lip protrusion, wider lip opening by more tongue retraction. This study investigates whether spe...
متن کاملVelar Movements for Two French Speakers
This study compares velar movements for nasal vowels and consonants; it investigates contextual nasalisation; and it provides new data on how nasalisation is affected by speech rate. Velar position is measured with an electromagnetic articulatograph (EMA) for two French speakers. Our results confirm that (i) nasal vowels are produced with a lower velum height than nasal consonants; (ii) the con...
متن کاملNonsegmental Influences on Velum Movement Patterns: Syllables, Sentences, Stress, and Speaking Rate*
Investigations of the motor organization of speech show that if we can identify individual segmental requirements, we can begin to predict the manner in which segments will influence each other in fluent speech. That is, we can model coarticulation as the outcome of temporal overlap (coproduction) among characteristic speech movements for successive segments (e.g., BellBerti & Harris, 1981; Fow...
متن کاملTowards Non-invasive Velum State Detection during Speaking Using High-frequency Acoustic Chirps
This paper presents our progress towards a convenient and non-invasive real-time method to measure the state of the velum (raised vs. lowered) that works both during normal and silent speaking. The method emits acoustic signals with a power band from 12 to 24 kHz into a nostril and analyzes the “echo” from the nasal cavity. Here we describe two design iterations of the method, present first tes...
متن کامل