FLEURS what it measures, how it is scored and who leads it

Name: FLEURS
Creator: Google
Published: 2022-05-25

Speech · Google · introduced May 25, 2022 · Word or character error rate

About this benchmark

Task: Recognise, identify and translate speech in 102 languages.
Dataset: Parallel read speech built on the FLoRes-101 sentences.
Method: Error rate for recognition, accuracy for language identification.
Metric: Word or character error rate
Organization: Google
Introduced: May 25, 2022
Paper: FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech (arXiv 2205.12446)
Home: huggingface.co/datasets/google/fleurs

Newest first. New versions are added, never rewritten.

Version	Date	What changed
1Released with the paper.	May 25, 20224 years ago	Released with the paper.

Sources: each benchmark's paper (arXiv) and its official site or leaderboard. Scores are the boards captured here every day, exactly as published.