HebDB: a Weakly Supervised Dataset for Hebrew Speech Processing
This addresses the problem of limited data for Hebrew speech processing, enabling better tools for Hebrew speakers, but it is incremental as it focuses on a specific language.
The authors tackled the lack of resources for Hebrew speech processing by introducing HebDB, a weakly supervised dataset with about 2500 hours of natural speech, and showed that their baseline ASR models optimized on this dataset outperform current multilingual alternatives with similar model sizes.
We present HebDB, a weakly supervised dataset for spoken language processing in the Hebrew language. HebDB offers roughly 2500 hours of natural and spontaneous speech recordings in the Hebrew language, consisting of a large variety of speakers and topics. We provide raw recordings together with a pre-processed, weakly supervised, and filtered version. The goal of HebDB is to further enhance research and development of spoken language processing tools for the Hebrew language. Hence, we additionally provide two baseline systems for Automatic Speech Recognition (ASR): (i) a self-supervised model; and (ii) a fully supervised model. We present the performance of these two methods optimized on HebDB and compare them to current multi-lingual ASR alternatives. Results suggest the proposed method reaches better results than the evaluated baselines considering similar model sizes. Dataset, code, and models are publicly available under https://pages.cs.huji.ac.il/adiyoss-lab/HebDB/.