1 paper · 1 filter
Zhennan Lin, Kaixun Huang, Wei Ren +2
Deep biasing improves automatic speech recognition (ASR) performance by incorporating contextual phrases. However, most existing methods enhance subwords in a contextual phrase as…