First and second-pass rescoring strategies can be leveraged together to improve the recognition of such words . We show that such an approach can improve personalized content recognition by up to 16% with minimum degradation on the general use case . We describe a fast and scalable algorithm that enables ourbiasing models to remain at the word-level, while applying the biasing at thesubword level . This has the advantage of not requiring the biased models to bedependently on any subword symbol table . We also describe a novel second passde-biasing approach: used in conjunction with a first-pass shallow fusion thatoptimizes on oracle WER, we can achieve an additional 14% improvement on personalizedcontent recognition, and even improve accuracy for the . general use

Author(s) : Aditya Gourav, Linda Liu, Ankur Gandhe, Yile Gu, Guitang Lan, Xiangyang Huang, Shashank Kalmane, Gautam Tiwari, Denis Filimonov, Ariya Rastrow, Andreas Stolcke, Ivan Bulyko

Links : PDF - Abstract

Code :
Coursera

Keywords : recognition - improve - level - approach - biasing -

Leave a Reply

Your email address will not be published. Required fields are marked *