Web Version of SlideSpeech

[John Graves]

At long last, the Open Allure Dialog System is moving onto the web, renamed SlideSpeech. The application continues to be Python-based, utilizing the CherryPy web framework on a server to deliver slide-presentation-to-web-presentation-with-voice-overs-and-interaction conversion. This should permit Wikipedia-style collaboration on presentations with visible slides (rather than merely via a wiki-based text-only script with links to slide images).

The current research server is running Ubuntu Linux and SlideSpeech version 1.4 does not yet have a working text-to-speech engine, but an excellent engine may soon be available from Cereproc. Meanwhile, please take a look at the latest authoring interface and the source code.

If you happen to create a test presentation, please post a link to it as a reply to this message.

Thanks.

[Grant Paton-Simpson]

Good work - sounds like it's all coming together.

[Dennis Daniels]

Where can we find demo videos /screencasts of same?

How do you use the research server UI?

[John Graves]

Thanks for the prompting Dennis. I've used the system on my Mac to produce this presentation about the new web-based authoring.

Again, the publicly available http://wikitospeech.pagekite.me research server has version 1.4 of the code and lacks a working text-to-speech engine, but these points should be addressed shortly (later today for version 1.6 of the code and hopefully next week for the Cereproc engine for Linux).

[John Graves]

OK, a live instance of web-based SlideSpeech authoring is now running on the research server, with code version 1.6, which features the ability to import an Open Document Presentation (.odp) format presentation and parse it into individual slide images (.png files) and the speaker notes for each slide. The speaker notes become text that anyone accessing the presentation on-line can modify. So this system now offers a working prototype of collaborative (one person at a time) work on a presentation.

Again, this is just using eSpeak (an old and very "computerish" voice), but that should change very soon.

To "clone" a presentation, change the script name and publish the clone.
Add interaction.
More tutorial videos coming soon ...