Registration Oct 7 — 13 Competition Oct 19 — 20 Judging Oct 21 — 24

Launch Site see me speak

This is an entry in the 2013 Rails Rumble. After checking it out, we encourage you
to leave feedback and then take a look at the hundreds of other amazing entries.

V200_team-39

Description

Filed in Education & Learning.

See Me Speak is a dictionary for sign language. You can easily search for signs and even add missing ones with your webcam.

We wanted to build something both useful and helpful: there are currently no easy ways to document and learn sign languages on the internet.

Technologicaly, we wanted to play with HTML 5 video capture (sadly only works properly in Chrome at the moment).

We'd love to have your support, because we still need a lot of words.

Built By Team "seemespeak"

Built With

Interested in seeing what was used to build this application? Resources may include things like open source libraries, third party APIs, and icons.

The website is a simple Rails frontend built on Twitter Bootstrap (http://getbootstrap.com/). For the Logo, we used the Gaudette sign language font: http://lifeprint.com/asl101/pages-layout/gallaudettruetypefont.htm. The s...

Read More...

The website is a simple Rails frontend built on Twitter Bootstrap (http://getbootstrap.com/). For the Logo, we used the Gaudette sign language font: http://lifeprint.com/asl101/pages-layout/gallaudettruetypefont.htm. The site is fully translated using Rails i18n in both english and german.

For seed content, we used the freely available corpus in german sign language available at: http://dgs.wikisign.org/. As we expect many of our visitors to be english-speaking, we translate the german words into english using Bing Translator: http://www.bing.com/translator/

As our backend, we use elasticsearch (http://www.elasticsearch.org/) exclusively. We use Virtus (http://github.com/solnic/virtus) to provide models on top of elasticsearch.

The web recording functions are a version of MediaStreamRecorder (https://github.com/streamproc/MediaStreamRecorder)

The whole website runs on torquebox: http://torquebox.org/. We use torquebox internal queuing mechanisms heavily.

The video conversion is built using libav: http://libav.org/.

We use Google Analytics for tracking.

Source code is open: https://github.com/railsrumble/r13-team-39/

comments powered by Disqus