Skip to content
Language-Research-TechnologyPublic

About

Simple browser-based interface for displaying audio recordings and associated transcripts

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

42 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 

Repository files navigation

README

Transcribed Audio Collection Explorer (TrACE) is a simple web-interface for exploring audio recordings and associated transcripts, built with pure HTML and JavaScript without any dependencies.

Interface

All the code for the interface is contained in TrACE_template.html. You can just open the file in any browser, as long it has access to the data as described below. You can also rename the file itself as needed.

Interface color and title can be changed by editing the values in the TEMPLATE VARIABLES section at the top of the file. Values which may be of the immediate interest:

  • pageHeader -- header at the top of the page
  • annNoteTier -- tierID associated with notes or remarks (as opposed to, for example, direct utterances by the speakers). Annotations in this tier will have a separate highlighting.
  • --primary-color -- color of the interface. Can be a valid color name or a HEX string. Interface elements will be coloured with shades of this color.

Data structure

The interface assumes the existence of three data sources:

  • Folder with audio files
  • speakers.js -- information about speakers
  • transcripts.js -- time-stamped transcripts

It is assumed that all of them are in the same folder with TrACE_template.html.

speakers

The file should contain a Java Script constant named speakers. The constant should hold a JSON list, where each element is a JSON dictionary holding info about a single speaker. Each speaker should have at least one field "Name". Any other fields can have arbitrary names.

See an example file in the examples folder.

transcripts

The file should contain a Java Script constant named transcripts. The constant should hold a JSON list, where each element is a JSON dictionary holding annotations for a single transcript. The dictionary should have the following fields:

  • filename -- path to the audio file
  • displayName -- display name for the audio in the list of recordings
  • speakers -- a list of speakers present in the recording. Names of the spakers should match the names in the speakers.js file.
  • annotations -- a list of dictionaries, where each entry corresponds to a single annotation and contains the following fields:
    • ANNOTATION_ID
    • TIER_ID (contains a speaker's name)
    • content -- annotation text
    • start -- start timestamp in milliseconds
    • end -- end timestamp in milliseconds

See an example file in the examples folder.

Header logos

Logos of your project, partners or funders can be shown under the page header. When preparing TrACE with make_trace, you can list them in a CSV file (--logos_file in the command line, or "Logos CSV file" in the graphical interface). Each row corresponds to one logo, shown from left to right:

image,text,url,title
logos/anu.png,,https://www.anu.edu.au/,Australian National University
,My Project,https://example.org/,
  • image -- path to the image file, relative to the CSV file. Supported formats: PNG, JPEG, GIF, SVG, WebP.
  • text -- text to show instead of an image, for logos without an image
  • url -- page to open when the logo is clicked (optional)
  • title -- text shown when hovering the mouse over the logo (optional)

Each logo needs either an image or a text. Image files are embedded into the HTML file, so they don't need to be copied together with it.

Credits

TrACE was created by Wolfgang Barth and Anton Malko, as part of LDaCA team at the Australian National University.

About

Simple browser-based interface for displaying audio recordings and associated transcripts

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages