Project: Flipside
Kaldi-based ASR experiments for the Filipino language
Corpus-building
I am looking for some volunteers for transcribing public domain Filipino data. Anyone interested, please send me an email at fmang at ieee dot org.
Recipes
The recipes are still under continuous cosmetic work. If you're interested in helping out, please contact: Federico Ang (fmang at ieee dot org)
Pretrained Models
Pre-trained models using Kaldi are available at: https://alphacephei.com/vosk/models
Closed-captioning Samples
To see the performance of our models, please check out the following demos: https://drive.google.com/drive/folders/1X35dRcUpNrU2ulbR58pGpYWfSOLq4_wY?usp=sharing