MediaSpeech is a media speech dataset (you might have guessed this) built with the purpose of testing Automated Speech Recognition (ASR) systems performance. The dataset consists of short speech segments automatically extracted from media videos available on YouTube and manually transcribed, with some pre- and post-processing. The dataset contains 10 hours of speech for each language provided. This release contains audio datasets in French, Arabic, Turkish and Spanish, and is a part of a larger private dataset.
Source: MediaSpeech: Multilanguage ASR Benchmark and Dataset
Variants: MediaSpeech
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Speech Recognition | Quartznet | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
| Speech Recognition | Wit | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
| Speech Recognition | Azure | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
| Speech Recognition | VOSK | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
| Speech Recognition | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 | |
| Speech Recognition | wav2vec | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
| Speech Recognition | Deepspeech | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
| Speech Recognition | Silero | MediaSpeech: Multilanguage ASR Benchmark and … | 2021-03-30 |
Recent papers with results on this dataset: