Differences
This shows you the differences between two versions of the page.
| Both sides previous revisionPrevious revisionNext revision | Previous revision |
| public:superradiohost:main [2008/01/02 13:09] – thorisson | public:superradiohost:main [2024/04/29 13:33] (current) – external edit 127.0.0.1 |
|---|
| |
| ===== ===== | ===== ===== |
| **Superradiohost** constitutes the first phase of the [[http://cadia.ru.is|CADIA]] [[Public:Superhumanoids:Main | Superhumanoids]] project. It aims to make [[Public:Superhumanoid One:Main|S1]], the first Superhumanoid in the Superhumanoids series, into a bona fide - but completely artificial - radio show personality. Our goal is to make SuperRadioHost better than the average human radio host, as judged by listeners. Several key components of the SuperRadioHost have been developed including a sophisticated turntaking system. In 2008 we plan to get a public national radio show for S1. The tentative name for the show is "RadioActive with SuperRadioHost". | **Superradiohost** constitutes the first phase of the [[http://cadia.ru.is|CADIA]] [[Public:Superhumanoids:Main | Superhumanoids]] project. It aims to make [[Public:Superhumanoid One:Main|S1]], the first Superhumanoid in the Superhumanoids series, into a bona fide - but completely artificial - radio show personality. Our goal is to make SuperRadioHost better than the average human radio host, as judged by listeners. Several key components have already been developed for the Host including a sophisticated turntaking system. At some appropriate point in the future we plan to get a public national radio show for S1. The tentative name for the show is "RadioActive with SuperRadioHost". |
| |
| **Methodology**. We are using [[http://www.mindmakers.org/projects/CDM | Constructionist Design Methodology]] throughout, and building the architecture in the [[http://www.mindmakers.org/projects/Psyclone | Psyclone]] framework. To achieve realtime we are using 12 networked desktop machines to do parallel processing including speech and prosody recognition, planning, and various other processing. Two addtional workhorse computers are used for input (collecting and pre-processing audio) and output (speech synthesis and other mouth-sound generation). | **Methodology**. We are using [[http://www.mindmakers.org/projects/CDM | Constructionist Design Methodology]] throughout, and building the architecture in the [[http://www.mindmakers.org/projects/Psyclone | Psyclone]] framework. To achieve realtime we are using 12 networked desktop machines to do parallel processing including speech and prosody recognition, planning, and various other processing. Two addtional workhorse computers are used for input (collecting and pre-processing audio) and output (speech synthesis and other mouth-sound generation). |
