G.Houbart - Viewpoints On Demand

Viewpoints

On Demand
My thesis work at the Media lab was sponsored by the News In the Future consortium. NIF's goal is to invent the technologies that will shape future media. As part of the Interactive Cinema group, an important part of the research was exploring interactive narrative in video for fiction, news or documentaries.
Viewpoints On Demand (VOD) focuses on the news and information side. "It was a Knowledge War" was produced using the VOD technology.

The issue
The interaction model used in VOD offers an original approach to interactive video: The viewer only adjusts a few settings or "content knobs" (to use the term coined by Nicholas Negroponte) and the engine generates a custom video. By doing so, we offer a powerful level of control and interaction instead of only presenting "clips", like CD-ROMs and the Web usually do.
By constructing videos on the fly, we take advantage of more than a 100 years of cinema and the powerful language developed by film editors, as opposed to popping isolated clips on the screen. This approach combines a low level of interaction and high level of control: the viewer sets some parameters then sits back and watches the result.

This is not a system in which the viewer navigates through information (even though it can combine both). The point is to allow the user to use high level concepts such as opinions and point of view by adjusting parameters controlling editing.
You might want to see more of the views expressed by a particular journalist or more of the views of the former Director of Information for the Pentagon on the selected issue. May be you had planned to spend two hours but changed your mind in the middle of the documentary, asking for a shorter version. All these adjustments increase control over content. Such scenario could be seen as a component of a "smart VCR". This is still a linear medium but taken to another level.

Model of Interaction for the Viewer

The viewer sets 3 constraints which are used by a software engine to produce a story in the form of a video sequence:

1 ) on the first screen a viewpoint is selected to be emphasized

2) on the second screen one of the themes (or a combination of themes) covered in the video database is (are) selected

3) a length is defined: the system uses an algorithm to "compress" or "decompress" in time a given story.

4) A video is generated based on these constraints

This raises important issues on the author's side like:

- How to preserve the journalist's perspective? What becomes his job and input?

- Are there applications for this type of technology beyond video reports?
Among possible answers:
- Certain perspectives could be shown on demand (the New York Times' view on some event or Le Monde's view (French newspaper), the viewer could switch back and forth.
- They could be - also on demand - juxtaposed to other viewpoints.

In some instances, there might be no journalistic perspective: a video database for tourism could be created to answer invividual requests for information. A video sequence assembled on the fly would come as an answer to suggest major highlights depending on the locations, the time of the years, or the sports selected by the viewer.

Home | Online TV | Content | Media Lab | Startups | Design |
-------------------------------------------------------------------
Copyright 1994 - Gilberte Houbart - MIT


Viewpoints On Demand	My thesis work at the Media lab was sponsored by the News In the Future consortium. NIF's goal is to invent the technologies that will shape future media. As part of the Interactive Cinema group, an important part of the research was exploring interactive narrative in video for fiction, news or documentaries. Viewpoints On Demand (VOD) focuses on the news and information side. "It was a Knowledge War" was produced using the VOD technology.

The issue The interaction model used in VOD offers an original approach to interactive video: The viewer only adjusts a few settings or "content knobs" (to use the term coined by Nicholas Negroponte) and the engine generates a custom video. By doing so, we offer a powerful level of control and interaction instead of only presenting "clips", like CD-ROMs and the Web usually do. By constructing videos on the fly, we take advantage of more than a 100 years of cinema and the powerful language developed by film editors, as opposed to popping isolated clips on the screen. This approach combines a low level of interaction and high level of control: the viewer sets some parameters then sits back and watches the result. This is not a system in which the viewer navigates through information (even though it can combine both). The point is to allow the user to use high level concepts such as opinions and point of view by adjusting parameters controlling editing. You might want to see more of the views expressed by a particular journalist or more of the views of the former Director of Information for the Pentagon on the selected issue. May be you had planned to spend two hours but changed your mind in the middle of the documentary, asking for a shorter version. All these adjustments increase control over content. Such scenario could be seen as a component of a "smart VCR". This is still a linear medium but taken to another level. Model of Interaction for the Viewer The viewer sets 3 constraints which are used by a software engine to produce a story in the form of a video sequence:
	1 ) on the first screen a viewpoint is selected to be emphasized
	2) on the second screen one of the themes (or a combination of themes) covered in the video database is (are) selected
	3) a length is defined: the system uses an algorithm to "compress" or "decompress" in time a given story.
	4) A video is generated based on these constraints
This raises important issues on the author's side like: - How to preserve the journalist's perspective? What becomes his job and input? - Are there applications for this type of technology beyond video reports? Among possible answers: - Certain perspectives could be shown on demand (the New York Times' view on some event or Le Monde's view (French newspaper), the viewer could switch back and forth. - They could be - also on demand - juxtaposed to other viewpoints. In some instances, there might be no journalistic perspective: a video database for tourism could be created to answer invividual requests for information. A video sequence assembled on the fly would come as an answer to suggest major highlights depending on the locations, the time of the years, or the sports selected by the viewer. Home \| Online TV \| Content \| Media Lab \| Startups \| Design \| ------------------------------------------------------------------- Copyright 1994 - Gilberte Houbart - MIT