BeatMaker - A prosody-based computational tool for L2 pronunciation teaching, forensic and sociophonetic comparison, and clinical purposes

BeatMaker can be used for pronunciation teaching, speaker comparison in forensic and sociophonetic areas, and clinical purposes.

Through prosodic analysis (rhythm and intonation) utilizing low-pass filtered and VoiceBeat audio, the tool provides objective acoustic profiling across four core domains:

Pronunciation Teaching

• Prosodic Modeling: Assessing L2 learners' intonation by comparing their f0 contours directly against native-speaker targets via low-pass filtered audio.

• Rhythmic Transfer: Visualizing  speech rhythm continuum (in the lexical domain) to help students reduce first-language prosodic interference.

• VoiceBeat Audios: Utilizing real-time melodic feedback to isolate suprasegmental target features without the distracting influence of segmental phonetic errors.

Forensic Speaker Comparison

• Intonation Profiling: Extracting robust, idiosyncratic pitch contours and fundamental frequency distributions that persist even in low-quality or intercepted acoustic data.

• Temporal Biomarkers: Comparing stable, individualistic speaker habits such as habitual pause placements, speech rate variations, and phrase-level macro/micro-rhythms.

• Channel-Resistant Metrics: Leveraging low-pass filtering to isolate suprasegmental markers that remain reliable across varying forensic telephone line distortions or codec variations.

Sociophonetic Variation

• Dialectal Rhythmology: Quantifying differences between syllable-timed and stress-timed language varieties using macroscopic rhythmic intervals.

• Intonational Typology: Mapping regional and social variations in question-asking patterns, phrase-final pitch movements, and group-specific prosodic contours.

• Macro-Prosodic Styles: Analyzing how speakers modulate pitch range and temporal pacing across different social interactions or stylistic contexts.

Clinical & Neuropsychiatric Research

• Trisomy 21: Quantifying macroeconomic speech rhythm variations, altered syllable duration, and macroscopic changes in fundamental frequency ($F_0$) contours caused by hypotonia.

• Autism Spectrum Disorder (ASD): Mapping atypical suprasegmental prosody, including hyper- or hypo-variable intonation contours and macro-prosodic variations in pitch range.

• Schizophrenia: Evaluating flat affect and vocal monotony through pitch range compression, alongside measuring pause structures, speech-to-silence ratios, and temporal rhythm degradation.

ORIGINAL ARTICLE

Silva Jr., L. (2023). BeatMaker: a computational system for foreign language pronunciation teaching based on speech prosody. RENOTE, 21(1), 341–352. https://doi.org/10.22456/1679-1916.134363

Contact:: leonidas.silvajr@uepb.edu.br