PulseAugur
LIVE 06:32:48
research · [2 sources] ·
0
research

AI model predicts stuttering events from audio, deploys on-device

Researchers have developed a new Convolutional Neural Network (CNN) model capable of predicting upcoming stuttering events from short audio clips. The 616K-parameter model, trained on the SEP-28k dataset, demonstrates a particular aptitude for identifying precursors to severe stuttering events like blocks and sound repetitions. Notably, the model can be deployed on-device, with efficient export formats and low latency demonstrated on various Apple devices. AI

Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →

IMPACT Enables on-device prediction of stuttering events, potentially aiding in real-time intervention systems.

RANK_REASON This is a research paper detailing a new model and its evaluation.

Read on arXiv cs.LG →

COVERAGE [2]

  1. arXiv cs.LG TIER_1 · Nazar Kozak ·

    Predicting Upcoming Stuttering Events from Three-Second Audio: Stratified Evaluation Reveals Severity-Selective Precursors, and the Model Deploys Fully On-Device

    arXiv:2604.27279v1 Announce Type: cross Abstract: Audio-based stuttering systems to date have been trained for detection -- what disfluency is present now -- leaving prediction, the capability needed for closed-loop intervention, unstudied at deployable scale. We train a 616K-par…

  2. Hugging Face Daily Papers TIER_1 ·

    Predicting Upcoming Stuttering Events from Three-Second Audio: Stratified Evaluation Reveals Severity-Selective Precursors, and the Model Deploys Fully On-Device

    Audio-based stuttering systems to date have been trained for detection -- what disfluency is present now -- leaving prediction, the capability needed for closed-loop intervention, unstudied at deployable scale. We train a 616K-parameter CNN on SEP-28k (Apple, 20,131 three-second …