Welcome to the Speech Recording Project for the Persian Dysarthric Speech Dataset (PDSDS)! The largest collection of speech data from dysarthric patients for Automatic Speech Recognition (ASR) and AI research.
Scroll down to learn more about the project.
Fa |
Individuals with Dysarthria Speech Disorder often become isolated even within their own families. The Persian Dysarthric Speech Recognition Research Project and the PDSDS dataset are designed to reduce this isolation.
Conversation is an everyday occurrence for most people.
However, it can be a major challenge for individuals with non-standard speech (Dysarthric speech).
This research project is being carried out as part of a doctoral dissertation.
The goal of this project is to create a platform for recognizing the speech of individuals with speech disorders (dysarthria) using AI technology and to establish a reference Persian speech dataset of dysarthric patients for ASR and AI research.
Volunteers record simple and practical phrases.
Researchers analyze the audio data and improve Automatic Speech Recognition (ASR) algorithms.
This "Dataset" is used in assistive applications for understanding the speech of people with non-standard speech (paralytic speech), Persian speech recognition, and the development of communication-aid systems.
The following conditions are necessary to participate in this research study:
After registering as a volunteer for this research study, you will use the online recording tool provided on this website to read the phrases displayed on the screen aloud. These recordings will then be reviewed by the AI researcher (Reza Afshinmehr) and used to train AI speech recognition software.
You must have access to a phone, tablet, laptop, or desktop computer with a microphone. It doesn't matter if the microphone is internal or a professional external one. To start, simply register, then click the red "Record Phrases" button. After reading each displayed phrase aloud and clearly (preferably in a quiet environment), press the red button again (Stop Recording) to finish that phrase and move to the next one (this process repeats until the last phrase).
If for any reason you cannot perform the recording operation yourself, you can ask a nurse or family member to assist you.
If your problem is technical and you receive an error message, first check your internet connection, then open the site using a major browser like Google Chrome. If the problem persists, you can contact us.
As mentioned earlier, this project is a research study in the form of a doctoral dissertation by Mr. Reza Afshinmehr, a PhD student in Artificial Intelligence. The goal of this dissertation is to produce knowledge and find a way to recognize the speech of dysarthric patients (such as ALS, MS, etc.) using speech processing and AI technology. From the achievements of this work, in addition to enhancing the knowledge of our country's scientific community, you and all dear patients like you can benefit and experience easier communication in society.
This research project is still in its early stages and at the level of a theory. As you know, every research project worldwide requires extensive investigation, researcher scrutiny, and numerous tests for researchers to prove their hypothesis and "produce science," and this project is no exception. After completing the research phases, the project output will be provided to participants for trial use.
Yes, your information is kept strictly confidential (according to doctor-patient confidentiality principles) in the system, and no natural or legal person will have access to this information. This data is only for use in the future research process. Our only way to communicate with you is the contact number you entered in the system. We appreciate your cooperation in advance.
