Primary competition visual

TechCabal Ewè Audio Translation Challenge

$1 000 USD
Challenge completed ~1 year ago
Classification
Automatic Speech Recognition
267 joined
80 active
Starti
Aug 26, 24
Closei
Sep 29, 24
Reveali
Oct 10, 24
User avatar
hark99
Self-employed
Mono Channel Train Audio files have fewer features than Test Audio Files
Data · 19 Sep 2024, 04:28 · 0

@Zindi The test data contains around 260 files, which have more features. should we at least add some lengthy audio files to the train data so that the model can be trained on them, and achieve almost the same accuracy on both validation and test data? At the moment, there is a huge difference between the model accuracy (without any overfitting) and the leaderboard results.

Discussion 0 answers