k2-fsa / k2

FSA/FST algorithms, differentiable, with PyTorch compatibility.
https://k2-fsa.github.io/k2
Apache License 2.0
1.11k stars 213 forks source link

Spoken Language Understanding with FSTs #1006

Open DanBmh opened 2 years ago

DanBmh commented 2 years ago

Hi, I would like to notify you about my project finstreder, which is combining FSTs with Intent and Entity tagging for Spoken Language Understanding. The approach itself is quite simple and fast, but could outperform most other algorithms on smaller SLU datasets. Maybe you can find something interesting there for your trainable FSTs. You can also find an overview of the approach in the paper about finstreder. Greetings Daniel

jtrmal commented 2 years ago

interesting, thanks for sharing

On Thu, Jul 7, 2022 at 6:54 AM DanBmh @.***> wrote:

Hi, I would like to notify you about my project finstreder https://gitlab.com/Jaco-Assistant/finstreder, which is combining FSTs with Intent and Entity tagging for Spoken Language Understanding. The approach itself is quite simple and fast, but could outperform most other algorithms on smaller SLU datasets. Maybe you can find something interesting there for your trainable FSTs. You can also find an overview of the approach in the paper https://arxiv.org/pdf/2206.14589.pdf about finstreder. Greetings Daniel

— Reply to this email directly, view it on GitHub https://github.com/k2-fsa/k2/issues/1006, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACUKYX36NBPEJEMZU2YV3RLVS2ZODANCNFSM52454AIA . You are receiving this because you are subscribed to this thread.Message ID: @.***>