Skip to content

clamsproject/app-nfa-wrapper

Repository files navigation

NFA Wrapper

Description

Wraps the NVIDIA NeMo Forced Aligner tool to temporally align transcribed text with its audio source.

User instructions

General user instructions for CLAMS apps are available at CLAMS Apps documentation.

System requirements

This app requires Python 3.10.12 or higher. For local installation of required Python modules, see requirements.txt.

Configurable runtime parameters

For the full list of parameters, please refer to the app metadata from the CLAMS App Directory or the metadata.py file in this repository.

Input and output details

This app accepts an empty MMIF file with the file locations of the required AudioDocument/VideoDocument and TextDocument sources.

Example input:

clams source audio:/path/to/source text:/path/to/text

The app outputs a Token annotation corresponding to each whitespace-delimited token in the source transcript, as well as a TimeFrame annotation identifying the audio segment where it appears and an Alignment annotation linking the Token and TimeFrame. For more details, see the output section of the app metadata.

About

Wrapper for Nemo forced aligner tools

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors 3

  •  
  •  
  •