Skip to content
Thoth Tech

OnTrack Voice Verification User Document

Author Information

  • Author: Ha Nguyen
  • Team: OnTrack - Voice Verification
  • Team (Delivery and/or Product) Lead: ShaeChristmas

Document Summary

  • Documentation Title: Voice Verification User Document
  • Documentation Type: Documentation
  • Documentation Information Summary: User Design Document detailing guide on Enrolment - How Students can register their voice to Voice Verification System

Document Review Information

  • Date of Original Document Submission to GitHub: 28/09/2022
  • Documentation Version: 1.0
  • Date of Previous Documentation Review: 28/09/2022
  • Date of Next Documentation Review: to be decided

Key Terms

A voiceprint is another way to use your unique features to identify who you are, similar to a fingerprint.

Key Links/Resources

Contacts for further information

See Thoth Tech Handbook.

Delivery Description

The OnTrack Voice Verification system allows student to enrol their voice and use their voice print to verify their identity when discussing or submitting works

This system would be implemented inside the existing OnTrack Project, and integrated into OnTrack by using the pre-existing audio submission system.

Voice Verification system has two phases:

  • Enrolment - Student’s voice is recorded and specific voice features are extracted into a voice print.

  • Verification - Student’s audio submission is compared against a previously created voice print.

Main Process

Proposed how voice verification system work

  • Verified: The audio file passed a certain confidence value and concluded as same person
  • Unverified:The audio file is under a confidence value range and concluded as not a same person
  • Pending: The audio is pending/in awaiting queue for comparing

Constraints

  1. Login to OnTrack site as an Deakin student
  2. Allow browser use device’s microphone
  3. Speaking language: English
  4. The voice must be between three seconds and one minute
  5. The volumes must not exceed 5 MB
  6. Supported file types: .wav, mp3, m4a, .flac (now the voice system only accepts .flac type files)

Tips: Speak at a normal cadence and clearly.

Detailed Process

Voice Enrolment Check Flow

Voice Enrolment GUI

Voice Enrolment GUI

Voice Enrolment GUI

References

Speaker Verification Enrolment Flow Example