Re: Joining the Media Fragment WG and contribution for the use cases of media fragments

Dear Raphael,

(My apologies that I can not spell your name correctly including 
"e + umlaut" in my environment...)

Thank you for your comments and warm welcome message.

| Thanks for your contribution. I have seen you're co-chairing the MAMI XG 
| and I wonder if you would like to act as our liaison with this XG?

Yes, I can act as your liaison with MAMI XG.

At this time, I'm not sure if I had better to act as your liaison with
MAMI XG or act as just one member of Media Fragments WG because the MAMI
XG has just launched last month.

I think the areas of interest of both Media Fragments WG and MAMI XG are
deeply related. I would like to find some collaboration points of both
groups together.

| We have briefly discussed your document in today's telecon. As Silvia 
| said, you bring more "water to our mill" in the sense that the face or 
| clothes recognition you describe corresponds to the requirements of our 
| scenario 2, 
| http://www.w3.org/2008/WebVideo/Fragments/WD-media-fragments-reqs/#scenario1.2 
| (Linking and Display use case).
| We could imagine adding your text as substitution for this use case.

Thank you very much for your discussion in the telecon yesterday.

I'm sorry that I could not attend the telecon yesterday. I'll try to
attend the telecons of Media Fragments WG from next week. I'm not sure
if I will be able to phone-in the telecons, but at least, I'll try to
join on IRC.

| We also raised a number of questions:
|    - do you need to specify 2 regions within a single URI? Currently, 
| the media fragment URI does not allow this use case
|    - do you need to specify moving regions within a media fragment URI?

My answers to both of your questions are "Yes".

In the trajectory analysis use case, we need to input several regions in
an image to the object coordinate locator engine.

In all of the 3 use cases I proposed, we need to specify moving regions
with a media fragment URI as a fragment correspond to an unique person.

There will be some more requirements for the media fragments URIs as I
mentioned in my reply to Silvia below:
-----
Considering above aspect, I think some new requirements for media
fragments URIs might be raised up. For example, how to embed the
accuracy of cropping in media fragments URIs, a negotiation protocols
for aligning accuracy of media fragments, exchanging combination of
several fragments in different images, etc.
-----

The details of above requirements are not clear enough in my mind yet.

By proposing my use cases, I wanted to clarify that there would be some
needs for Media Fragments URIs to apply them as "the analysis data
exchange interfaces among various media analysis engines in server-side".

I want my proposal will be a starting point for the discussions of these
furthur requirements for (the next?) media fragments URIs specifications.

BTW, can I create a new page on Media Fragments WG wiki[1] and put my
proposals on the page for further discussions and updates of my use
cases proposal?
[1] http://www.w3.org/2008/WebVideo/Fragments/wiki/Main_Page

With my best regards,
Nobu.

-----
Nobuhisa Shiraishi <7M3RLA/1>
Information and Media Processing Laboratories, NEC Corporation.
E-mail:n-shiraishi@bq.jp.nec.com Tel:+81-44-431-7672 Fax:+81-44-431-7588

Received on Thursday, 13 January 2011 04:46:17 UTC