Finals Act2 Tinimbang Santiago

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 4

San Sebastian College – Recoletos de Cavite

College of Engineering and Computer Studies


Embedded System 3

Finals. Activity 2
Voice Authentication App

TINIMBANG, JERICHO CARLO A.


SANTIAGO, REI FRANCIS A
4th year BSCPE
25 October 23

Paolo Roberto O. Lozada, Meng-CpE


Instructor
OBJECTIVES
 To train an audio classifier from scratch to distinguish between two voices.
 To use the PersonalAudioClassifier extension to authenticate users based on
their voices.
 To create a Voice Authentication Diary App that allows only authorized users
to access their diaries.
 To build an app using MIT App Inventor and integrate voice authentication
to enhance security.

MATERIALS, TOOLS AND EQUIPMENT


 A computer with a compatible web browser.
 Microphone for recording voice clips.
 Access to the https://c1.appinventor.mit.edu/ website.
 A friend or family member to serve as a partner for voice training.
 The Voice Authentication Diary App Inventor project file (aia) provided in the
instructions.
 MIT App Inventor platform for app development.
 Mobile device or emulator for testing the app.
 PersonalAudioClassifier extension for voice classification.
 TextToSpeech component for providing feedback.
 Internet connection for uploading the model.

DISCUSSION
In this project, you are tasked with creating a Voice Authentication App
using MIT App Inventor and the PersonalAudioClassifier extension. The primary
focus is on training an audio classifier to distinguish between two voices, which will
be used for voice-based authentication in a diary app. The discussion should cover
the following key points:

 The importance of voice authentication in enhancing app security.


 The process of training an audio classifier by recording voice samples from
two individuals.
 Uploading the trained model to the app using the PersonalAudioClassifier
extension.

 Implementing the logic to allow access to the diary app only if the voice
matches the authorized user.
 The use of TextToSpeech to provide feedback to users based on the voice
classification results.

OBSERVATION

Based on our observations, we have noticed that when we repeatedly train


the primary user, Vandan, by conducting the process multiple times (as we did by
attempting it 25 times), and also incorporating a training phase for another
individual, we are confident in our ability to detect the primary user, Vandan, with
approximately 89% certainty.

CONCLUSION
In summary, the primary objective of this project was to develop a Voice
Authentication App using MIT App Inventor and the PersonalAudioClassifier
extension. This app was designed to enhance the security of a diary application by
training the audio classifier to distinguish between two distinct voices, thereby
ensuring that only authorized users could access their diaries.

Throughout the course of this project, you have acquired valuable insights into the
processes of training an audio classifier, integrating it into an application, and using
TextToSpeech for user feedback. By diligently following the provided guidelines,
you have successfully constructed a diary application with voice authentication
capabilities, significantly enhancing the protection of personal content.

In a broader context, this project serves as a practical demonstration of the real-


world application of voice recognition technology for authentication purposes.
Furthermore, it lays the groundwork for the future development of more secure
and customized applications.

You might also like