Experiments on Turkish ASR with Self-Supervised Speech Representation Learning
This work addresses ASR for the low-resource Turkish language, but it is incremental as it applies an existing method to new data without achieving commercial viability.
The researchers tackled Turkish automatic speech recognition by pre-training HUBERT on 6,500 hours of YouTube speech data, but found the models were not robust enough for commercial use due to issues with accents, noise, and other real-world disturbances.
While the Turkish language is listed among low-resource languages, literature on Turkish automatic speech recognition (ASR) is relatively old. In this report, we present our findings on Turkish ASR with speech representation learning using HUBERT. We investigate pre-training HUBERT for Turkish with large-scale data curated from online resources. We pre-train our model using 6,500 hours of speech data from YouTube. The results show that the models are not ready for commercial use since they are not robust against disturbances that typically occur in real-world settings such as variations in accents, slang, background noise and interference. We analyze typical errors and the limitations of the models for use in commercial settings.