HIGHLIGHTS
- who: Guangzhi Sun and collaborators from the (UNIVERSITY) have published the paper: Minimising Biasing Word Errors for Contextual ASR with the Tree-Constrained Pointer Generator, in the Journal: (JOURNAL)
- what: In this paper, a tree-constrained pointer generator (TCPGen) component is proposed and developed for end-to-end contextual speech recognition. As it is hard to train a randomly initialised model with the MB the authors loss, which is similar to MWE, the MB the authors loss is applied from the epoch when the model is reasonably trained with CE loss, which depends on optimisation . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.