Libri-Light
Introduced by Kahn et al. in Libri-Light: A Benchmark for ASR with Limited or No Supervision
Libri-Light is a collection of spoken English audio suitable for training speech recognition systems under limited or no supervision. It is derived from open-source audio books from the LibriVox project. It contains over 60K hours of audio.
Source: Libri-Light: A Benchmark for ASR with Limited or No SupervisionBenchmarks
| Trend | Task | Dataset Variant | Best Model | Paper | Code |
|---|---|---|---|---|---|
|
Libri-Light test-clean
|
wav2vec 2.0 Large-10h-LV-60k
|
|||
|
Libri-Light test-other
|
wav2vec 2.0 Large-10h-LV-60k
|
Papers
| Paper | Code | Results | Date | Stars |
|---|
