TechCabal Ewè Audio Translation Challenge 💊

TechCabal Ewè Audio Translation Challenge

$1 000 USD

Challenge completed ~1 year ago

Skills you will learn

Classification

Automatic Speech Recognition

267 joined

80 active

Info Data Chat Leaderboard

Start

Aug 26, 24

Sep 29, 24

Reveal

Oct 10, 24

About

There are 5334 audio files in train and 2286 audio files in test.

The dataset was built to ensure equal representation of men and women. Data augmentation was used to generate more audio using initial contributions. Initially released in April 2024 with 800+ recordings, the resulting dataset currently has 7000+ entries of Ewe speakers, speaking in the following contexts :

Crossing road background noise
Rain and thunder background noise
Rural and forest background noise
Firefighters alarm background noise

The dataset was built 100% by Umbaji community members with the support of Google Cloud For Startups in 2023, as well as Microsoft Entrepreneurship For Positive Impact.

The audio files are available in this Google Folder.

Additional test audio files are available in this Google Folder.

Files

Description

Files

Train contains the target. This is the dataset that you will use to train your model.

Updated test file that contains more testing files. Test resembles Train.csv but without the target-related columns. This is the dataset on which you will apply your model to.

Updated ss, this file contains the IDs for the additional testing files provided in Test_1. Is an example of what your submission file should look like. The order of the rows does not matter, but the names of the "ID" must be correct.

Join the largest network for
data scientists and AI builders

About FAQs

Status