KT-Speech-Crawler: Automatic Dataset Construction for Speech Recognition from YouTube Videos

In this paper, we describe KT-Speech-Crawler: an approach for automatic dataset construction for speech recognition by crawling YouTube videos. We outline several filtering and postprocessing steps, which extract samples that can be used for training end-to-end neural speech recognition systems. In our experiments, we demonstrate that a single-core version of the crawler can obtain around 150 hours of transcribed speech within a day, containing an estimated 3.5% word error rate in the transcriptions. Automatically collected samples contain reading and spontaneous speech recorded in various conditions including background noise and music, distant microphone recordings, and a variety of accents and reverberation. When training a deep neural network on speech recognition, we observed around 40% word error rate reduction on the Wall Street Journal dataset by integrating 200 hours of the collected samples into the training set. The demo1 and the crawler code2 are publicly available.

@InProceedings{LWMW18, 
 	 author =  {Lakomkin, Egor and Weber, Cornelius and Magg, Sven and Wermter, Stefan},  
 	 title = {KT-Speech-Crawler: Automatic Dataset Construction for Speech Recognition from YouTube Videos}, 
 	 booktitle = {Proceedings of the Conference on Empirical Methods in Natural Language Processing},
 	 editors = {},
 	 number = {},
 	 volume = {},
 	 pages = {90--95},
 	 year = {2018},
 	 month = {Jan},
 	 publisher = {Association for Computational Lingustics},
 	 doi = {10.18653/v1/D18-2016}, 
 }