CW Pakistan
  • Legacy
    • Legacy Editorial
    • Editor’s Note
  • Academy
  • Wired
  • Cellcos
  • PayTech
  • Business
  • Ignite
  • Digital Pakistan
  • DFDI
  • PSEB
  • PASHA
  • TechAdvisor
  • GamePro
  • Partnerships
  • PCWorld
  • Macworld
  • Infoworld
  • TechHive
  • TechAdvisor
0
0
0
0
0
Subscribe
CW Pakistan
CW Pakistan CW Pakistan
  • Legacy
    • Legacy Editorial
    • Editor’s Note
  • Academy
  • Wired
  • Cellcos
  • PayTech
  • Business
  • Ignite
  • Digital Pakistan
  • DFDI
  • PSEB
  • PASHA
  • TechAdvisor
  • GamePro
  • Partnerships
  • Wired

Microsoft AI Powered To Clone Voice By 3-Second Audio

  • January 16, 2023
Total
0
Shares
0
0
0
Share
Tweet
Share
Share
Share
Share

With just a three-second audio clip, Microsoft’s new text-to-speech AI will be able to duplicate voices, including tone and pitch.


Despite being a complicated system, VALL-“neural E’s codec language model” is extremely simple to use and only requires the insertion of audio and text. The developers of the programme are certain that it can be applied to high-quality text-to-speech tasks including speech modifying and audio content production. Microsoft’s application is based on EnCodec, which Meta unveiled in October of the previous year.


VALL-E analyses how someone sounds and separates that information into discrete components, producing discrete audio codec codes from text and acoustic stimuli. EnCodec compares what it knows about how that voice would sound if it delivered a different phrase to training data.

The speech-synthesis abilities of VALL-E were taught using audio from a library that Meta put together, which contained 60,000 hours of English speakers from more than 7,000 speakers. The submitted training data must closely match the three-second voice clip sample for a successful outcome.

Although speech-generating software is frequently used by news sites, it requires a lot of input. What’s more, the voice lacks a human-like quality and is unable to communicate expressions or inflections. The application carries possible problems in the misuse of the model, such as spoofing voice recognition or mimicking a certain speaker. VALL-E is extremely advanced and offers a better and more accurate result with minimum required input.

By altering the random seed used in the generating process, the computer can produce differences in voice tone, as shown by the sample provided by Microsoft. VALL-E can simulate the acoustic environment of the audio that was present in the sample audio, such as simulating a voice over the phone.

Share
Tweet
Share
Share
Share
Previous Article
  • Business
  • Wired

SBP Easing Off IT Exporters To Increase Their Income

  • January 15, 2023
Read More
Next Article
  • Cellcos

Pakistan Stands At Second Highest Average Mobile Broadband Speed: PTA

  • January 16, 2023
Read More
You May Also Like
Read More
  • Wired

OIC COMSTECH Workshop Emphasizes Quality Assurance and University Rankings

  • Press Desk
  • July 11, 2025
Read More
  • Wired

Islamabad Court Suspends Ban on 27 Pakistani YouTube Channels

  • Press Desk
  • July 11, 2025
Read More
  • Wired

Pakistan and China Strengthen Media Ties to Counter Disinformation and Enhance Content Exchange

  • Press Desk
  • July 10, 2025
Read More
  • Wired

NCCIA Raids Illegal Faisalabad Call Centre Running Ponzi Scheme, 149 Arrested

  • Press Desk
  • July 9, 2025
Read More
  • Wired

KP Police Deploy Drone-Jamming Tech During Ashura Processions in Peshawar

  • Press Desk
  • July 7, 2025
Read More
  • Wired

CM Punjab Imposes Mobile Phone Ban on Doctors and Nurses in Public Hospitals

  • Press Desk
  • July 6, 2025
Read More
  • Wired

Microsoft’s Operational Restructuring: Ensure Continued Commitment to Pakistan

  • webdesk
  • July 3, 2025
Read More
  • Wired

Young Pakistani Scientists Represent Country at 74th Lindau Nobel Laureate Meeting

  • Press Desk
  • July 3, 2025

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Trending Posts
  • KP Launches Digital NOC System to Simplify Travel for Foreign Tourists
    • July 12, 2025
  • Shaza Fatima, Alibaba Discuss Strengthening Pakistan’s Digital Trade and Global E-Commerce Ties
    • July 12, 2025
  • Pakistan Delays Satellite Internet Launch to Finalize Rules, Attract More LEO Operators
    • July 12, 2025
  • IESCO Launches Self-Meter Reading Feature Through Apna Meter Apni Reading App
    • July 12, 2025
  • Pakistan Railways to Launch Digital Business Train with Wi-Fi, Enhanced Onboard Services
    • July 12, 2025
about
CWPK Legacy
Launched in 1967 internationally, ComputerWorld is the oldest tech magazine/media property in the world. In Pakistan, ComputerWorld was launched in 1995. Initially providing news to IT executives only, once CIO Pakistan, its sister brand from the same family, was launched and took over the enterprise reporting domain in Pakistan, CWPK has emerged as a holistic technology media platform reporting everything tech in the country. It remains the oldest continuous IT publishing brand in the country and in 2025 is set to turn 30 years old, which will be its biggest benchmark and a legacy it hopes to continue for years to come. CWPK is part of the SPIN/IDG Wakhan media umbrella.
Read more
Explore Computerworld Sites Globally
  • computerworld.es
  • computerworld.com.pt
  • computerworld.com
  • cw.no
  • computerworldmexico.com.mx
  • computerwoche.de
  • computersweden.idg.se
  • computerworld.hu
Content from other IDG brands
  • PCWorld
  • Macworld
  • Infoworld
  • TechHive
  • TechAdvisor
CW Pakistan CW Pakistan
  • CWPK
  • CXO
  • DEMO
  • WALLET

CW Media & all its sub-brands are copyrighted to SPIN-IDG Wakhan Media Inc., the publishing arm of NCC-RP Group. This site is designed by Crunch Collective. ©️1995-2025. Read Privacy Policy.

Input your search keywords and press Enter.