Skip to content

Developed and trained Gated-CNN models to detect types of stutter in speech and SVM classifier to suggest new therapies to the user according to his stutter type and severity

mansi-k/Stutter-Therapy

Repository files navigation

Stutter Diagnosis and Therapy System based on Deep Learning:

https://arxiv.org/abs/2007.08003

This work attempts to detect and classify stutters in the input audio while existing works focus only on detection of stutters. We have used Gated Recurrent CNN on MFCC audio features for stutter detection and classification. We propose an SVM-based system that can suggest therapies based on the type and severity of the stuttering. We have also developed an Android app, Node.js based server and exploited Firebase for storing API requests and responses.

Link to the pre-processed dataset: link

About

Developed and trained Gated-CNN models to detect types of stutter in speech and SVM classifier to suggest new therapies to the user according to his stutter type and severity

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published