This model is a fine-tuned version of Wav2Vec2-XLSR-53. We finetuned this model on the MGB-3 and Egyptian Arabic Conversational Speech Corpus datasets, acheiving WER of 29.3755%.

The performance of the model on the datasets is the following:

Model fine-tuning and data processing for this work were performed as a part of a Graduation Project from Faculty of Engineering, Alexandria University, CCE Program.