4 Comments

Summary:

A new project dubbed Universal Subtitles aims to launch a Wikipedia for subtitles and tools that make it easy for volunteers to add closed captions to any video they find online. Universal Subtitles is a new project by the Participatory Culture Foundation (PCF), better known as […]

A new project dubbed Universal Subtitles aims to launch a Wikipedia for subtitles and tools that make it easy for volunteers to add closed captions to any video they find online. Universal Subtitles is a new project by the Participatory Culture Foundation (PCF), better known as the makers of the Miro video player.

PCF co-founder Nicholas Reville told me that the group plans to release a first public demo of its tools in four to six weeks. He said that the group received an initial grant from the Mozilla Foundation to build Universal Subtitles, and that it is actively looking for volunteers through Mozilla’s new Drumbeat community.

The PCF has been building a subtitle widget that will allow volunteers to transcribe videos from pretty much any web site, sync them with the video in question and then publish the results online. Universalsubtitles.org will host these subtitles in a Wikipedia-like fashion, complete with different versions for each video and various ways for people to collaborate.

Subtitles will be saved in a format like .srt, which is supported by many video players. Reville told me that the site wants to enhance this with a new protocol that can be used by video players, web sites and various devices to query the Universal Subtitles site or any other subtitle repository and download the correct subtitle in the appropriate format. “We want to work with other organizations to create a new open standard that anyone who hosts subtitles or make a video player can use,” he said.

Universal Subtitles isn’t the only one working on enhancing web video through subtitles. YouTube recently rolled out site-wide closed captions through voice recognition. Reville said that Universal Subtitles can be complementary to YouTube’s efforts. “There are lots of things that speech recognition and machine translation can’t do well now,” he explained, adding that his group is also looking at options to incorporate voice recognition to make human subtitling even easier.

Related content on GigaOm Pro: Connected Consumer Market Overview, Q1 2010 (subscription required)

You’re subscribed! If you like, you can update your settings

  1. Thanks for the post, Janko!

    Also, maybe getting overly technical here, we’ll be agnostic towards the format of the subtitles (.srt, timed text, etc). Instead we’re looking towards a standard look-up protocol that any client can request any format of subtitles from a server.

    It’ll work like CDDB does for audio CD’s, only less centralized.

    1. Janko Roettgers Dean Tuesday, April 13, 2010

      Dean, thanks for the clarification!

  2. YouTube started a new project and had captioning companies compete because the voice recognition wasn’t accurate. http://youtube-global.blogspot.com/2010/06/professional-caption-services-get.html

  3. New Bill to Mandate Captions for Web TV Wednesday, August 4, 2010

    [...] and others if you want more accuracy for your transcripts. The Participatory Culture Foundation recently unveiled a project to crowdsource the transcription of web [...]

Comments have been disabled for this post