An initiative to teach Uzbek speech to computers

You and I can do it! Contribute to popularizing the Uzbek language!

What is the goal?

To help developers, researchers, startups - by providing an open and free Uzbek speech database.

goal-1

Research

Creating opportunities for scientific cooperation and research in the field of artificial intelligence
goal-1

Creativity

Opening the way for creativity to developers in the field of artificial intelligence
goal-1

Public participation

Implementation of our actions with public participation
goal-1

Open Source

Collecting the open-source Uzbek speech database
goal-1

Research

Creating opportunities for scientific cooperation and research in the field of artificial intelligence
goal-1

Creativity

Opening the way for creativity to developers in the field of artificial intelligence
goal-1

Public participation

Implementation of our actions with public participation
goal-1

Open Source

Collecting the open-source Uzbek speech database
goal-1

Research

Creating opportunities for scientific cooperation and research in the field of artificial intelligence
goal-1

Creativity

Opening the way for creativity to developers in the field of artificial intelligence

Voice collection campaign phases

1-stage

finished
200+

More than 200 hours of audio data were collected

1 October 2021 - 15 December 2021

2-stage

finished
1000+

More than 1000 hours of audio data were collected

21 October 2022 - 13 November 2022

3-stage

soon
3000+

At this stage, we want to collect more than 3000 audio data

...


How can I contribute?

Microphone

Read the sentence out loud

Record your voice by reading the displayed text in Uzbek on the Mozilla Common Voice platform. This is one of the most fun activities!

Microphone

Listen and check the recorded voice

Compare if the recorded sentence matches the text. It's very simple.

Microphone

Verify the texts

Check the texts based on grammar rules. The verified sentence is then added to the database.

Microphone

Read the sentence out loud

Record your voice by reading the displayed text in Uzbek on the Mozilla Common Voice platform. This is one of the most fun activities!

Microphone

Listen and check the recorded voice

Compare if the recorded sentence matches the text. It's very simple.

Microphone

Verify the texts

Check the texts based on grammar rules. The verified sentence is then added to the database.

Microphone

Read the sentence out loud

Record your voice by reading the displayed text in Uzbek on the Mozilla Common Voice platform. This is one of the most fun activities!

Microphone

Listen and check the recorded voice

Compare if the recorded sentence matches the text. It's very simple.

Microphone

Verify the texts

Check the texts based on grammar rules. The verified sentence is then added to the database.

Leading members

The most active listener

Leader
Xolmatov Shokirbek
48987
Leader
Mamatqulova Anora
47368
Leader
Darxonov Sherzod
33966

The most active speaker

Leader
Muhammaddiyor Teshaboyev
32373.53
Leader
Furqat Shahriddin
31501.41
Leader
Tojiyev Baxtiyor
29392.55

Cooperation with us

We are always open to cooperation with those who want to contribute to the development of artificial intelligence in Uzbekistan.

Read more
robot

Initiative partners

The UzbekVoice.ai team and partners unite enthusiastic people and invite like-minded people to corporate, institutional and scientific cooperation.

Uzbekvoice Team

partners1

Hasanov Bakhtiyor

Founder of UzbekVoice.ai

partners1

Abulfayz Mamatkulov

Frontend Developer

partners1

Sherbaev Dovron

Full Stack Developer

partners1

Muhammad Amin Kodirov

IT Project manager

partners1

Abdullayev Sherzod Shavkatjonovich

AI Researcher

partners1

Elmurod Kuriyozov

PhD in Computational Linguistics

partners1

Bekzod Buyukov

Backend Developer

partners1

Askar Abdullayev

Software engineer

partners1

Shakhriyor Khudoyberdiev

Backend developer

partners1

Liniza Fatkulina

Project Coordinator

partners1

Shohjahon Dilmurodov

Frontend Developer

partners1

Ramil Kamalov

Front-end developer

FAQ

A voice assistant needs enough audio data to understand a human voice.
Yes, of course! Because each region has its own dialect and pronunciation, we take this into account during the registration process and store it in the database.
Yes, of course! All our codes (bot, website and backend) are open source, the collected database is also open source.
1 point is given for voice recording and checking.
Even if he wins in two categories, one person will get one prize!