Automatic speech recognition
Datasets
| Name | License | Size | Date | Metadata |
|---|---|---|---|---|
| King Saud University Arabic Speech Database | Only research, no commercial | 590h | February 17, 2014 | Sampling rate: 48K, microphone speech, 269 speakers |
Papers
| Title | Summary | Date | Publisher |
|---|---|---|---|
| An enhanced automatic speech recognition system for Arabic | Develop an ASR system for MSA by using Kaldi toolkit, achieving WER: 12 | 2017 | Loria |
Frameworks
| Name | Publisher | Techs | Comments |
|---|---|---|---|
| OpenSeq2Seq | Nvidia | Python | End-to-end python pipeline, English pretrained models available |
| ESPnet | Researchers | Python,cpp, Kaldi | Features written on Disk. Joint CTC-Attention architecture |
More resources coming soon stay tuned ! 🤩 You are welcome to contribute to this project ! 🙏