W3C home > Mailing lists > Public > public-media-annotation@w3.org > November 2008

Re: [new use case suggestion] Use Case - Digital imaging lifecycle

From: Rubťn Tous <rtous@ac.upc.edu>
Date: Tue, 4 Nov 2008 18:40:37 +0100
Message-ID: <A900A673C14B404EBA4B59D91177D512@Vayito>
To: <public-media-annotation@w3.org>

Dear Werner, all,

thanks to all for your comments. I see a certain consensus regarding the 
need of a limitation in the complexity of the processing history metadata 
(Werner, Pierre-Antoine, Felix and Veronique) and I also agree with you 
because the goal of the group is multimedia metadata interoperability, not 
modelling. I also see that some of you agree in the need of having at least 
the possibility (not the obligation as pointed by Pierre-Antoine) to keep 
some information (Werner talks about the contributors, Veronique and Victor 
talk about separating idea/work/instance). If you agree, as suggested by 
RaphaŽl, I will add a tentative Use Case in the Wiki with a revised version 
of my initial idea in order to help the discussion. Pierre-Antoine suggested 
generalizing this issue to any kind of media, but maybe we can keep the use 
case as is for the discussion and later, if any new requirement arises, 
generalize the requirement.

Working in the use case a funny example came to my mind. Try to find in your 
favourite images search engine (e.g. Google Imanges) the image of a tattoo 
inspired in a famous painting (without using the name of a specific 
painting).

Best regards,

Ruben


----- Original Message ----- 
From: "Bailer, Werner" <werner.bailer@joanneum.at>
To: <vmalaise@few.vu.nl>; "Ruben Tous" <rtous@ac.upc.edu>
Cc: <public-media-annotation@w3.org>
Sent: Tuesday, November 04, 2008 5:50 PM
Subject: AW: [new use case suggestion] Use Case - Digital imaging lifecycle



Dear Ruben, Veronique, all,

I'm not sure about the inclusion "historic metadata" (actual I would suggest 
to use the term processing metadata or processing history metadata instead): 
This information can be quite detailed, specific to tools applied (e.g. 
settings, parameters), and is quite low-level information.

The set of people/organisations contributing to the creation of the media 
item (P_Meta uses the term "contributor", as this might involve less 
creative contributors such as movie producers) is relevant and I support 
Veronique's proposal. If necessary the type of contribution could be quite 
fine-grained in a certain application, without hindering other applications 
to deal with the concept of a generic creator.

Best regards,
Werner

-----UrsprŁngliche Nachricht-----
Von: public-media-annotation-request@w3.org im Auftrag von 
vmalaise@few.vu.nl
Gesendet: Di 04.11.2008 11:34
An: Felix Sasaki
Cc: Rubén Tous; public-media-annotation@w3.org
Betreff: Re: [new use case suggestion] Use Case - Digital imaging lifecycle


Quoting Felix Sasaki <fsasaki@w3.org>:

Dear all,

I also do think that our ontology should not go in too many details, but 
allow
placeholders for other schemas to fit in: having one generic "creator"
property/tag, and giving the possibility to scheme that are more refined to
extend this single property/tag into "conceptual creator" and "concrete
realisator of the piece". On the other hand, if we go for a description 
model
that keeps the distinction between the idea (the idea of a movie for 
example), a
realisation (one adaptation by a director) of the work and instances (a 
video
tape/DVD), it is possible to attach a property/tag of "creator" at all these
levels. the seamtic would be the agregation between the level of description
(idea/work/instance) and the role (creator).
But of course, this is just an idea, open to criticism... or approval :)
What do you think?

Best,
Veronique

>
> Hello Ruben, all,
>
> as you said below, in some formats like EXIF there is no separation
> between "historic" and metadata of the resource, and in others there is.
> Again I think we need to decide: how many details do we want to take
> into account? I think for metadata interoperability, the EXIF+others
> approach from the metadata WG is sufficient. What do you think?
>
> Felix


>
>
> Rub√©n Tous „.„,"„Įś>ł„„ĺ„-„Y:
> >
> > Dear all,
> >
> > it make sense to me to cover all the three main media categories
> > (video, still images and audio) as a hole or as three separated parts.
> >
> > However, the intention of my example was not so ambitious, it was just
> > related to what in DIG35 (cited in the PhotoUC) is named "History
> > Metadata":
> >
> >> From Section 3.2.4 in DIG35
> > (http://xml.coverpages.org/FU-Berlin-DIG35-v10-Sept00.pdf) :
> >
> > "For example, history may include certain processing steps that have
> > been applied to an image. Another example of a history would be the
> > image creation events including digital capture, exposure of negative
> > or reversal films, creation of prints, transmissive scans of negatives
> > or positive film, or reflective scans of prints. All of this metadata
> > is important for some applications. To permit flexibility in
> > construction of the image history metadata, two alternate
> > representations of the history are permitted"
> >
> > I think that EXIF and other formats mix this concept with the metadata
> > of the resource (e.g. the Exposure Time field in EXIF) but others like
> > DIG35 or MXF and AAF (Part 15 of
> > http://www.aafassociation.org/html/specs/aafobjectspec-v1.1.pdf talks
> > about Physical Essence) make a clear differentiation.
> >
> > What about a "History Metadata" Use Case?
> >
> > Best regards,
> >
> > Ruben
> >
> > ----- Original Message ----- From: <vmalaise@few.vu.nl>
> > To: "V√f¬≠ctor Rodr√f¬≠guez Doncel" <victorr@ac.upc.edu>
> > Cc: "Felix Sasaki" <fsasaki@w3.org>; "Pierre-Antoine Champin"
> > <swlists-040405@champin.net>; "Rub√f¬©n Tous" <rtous@ac.upc.edu>;
> > <public-media-annotation@w3.org>
> > Sent: Tuesday, November 04, 2008 9:58 AM
> > Subject: Re: [new use case suggestion] Use Case - Digital imaging
> > lifecycle
> >
> >
> >>
> >> Dear all,
> >>
> >> How about this solution: we could group a number of use cases under
> >> the "media"
> >> category, as we already have an "audio" use case, and take into
> >> account in the
> >> ontology 1.0 only the requirements that overlap with others? The
> >> description of
> >> the use case would show what other aspects still need to be taken into
> >> consideration when aiming for still images description compatibility.
> >>
> >> Best,
> >> Veronique
> >>
> >> Quoting V√f¬≠ctor Rodr√f¬≠guez Doncel <victorr@ac.upc.edu>:
> >>
> >>>
> >>> Hello all,
> >>>
> >>> I think it should be distinguished between the user roles regarding 
> >>> the
> >>> resource, and the user roles regarding the represented object.
> >>> Thus, the three kind of applications or roles defined by the
> >>> metadataworkinggroup (creator/changer/consumer) operate on the 
> >>> resource
> >>> but may not match logically the role regarding the represented object.
> >>>
> >>> For example, the word "creator" is somewhat ambiguous because it may
> >>> refer to the role which creates materially the resource, or to the
> >>> actual artist which conceives an idea. Both "creators" do not
> >>> necessarily match. Have you thought about it?
> >>>
> >>> Regards,
> >>> V√f¬≠ctor Rodr√f¬≠guez Doncel
> >>>
> >>>
> >>> Felix Sasaki escribi√f¬≥:
> >>> >
> >>> > Pierre-Antoine Champin √£¬.√£,"√£¬¬Į√¶>¬ł√£¬¬√£¬¬ĺ√£¬-√£¬Y:
> >>> >> Felix Sasaki a √f¬©crit :
> >>> >>>
> >>> >>> Hello Ruben, all,
> >>> >>>
> >>> >>> sorry for the late reply. Reading your proposal I think it is
> >>> >>> interesting for the photo use case. However I remember that we
> >>> >>> discussed at the f2f meeting about the focus of the Working Group,
> >>> >>> and most of the people want it to be video, with the possibility 
> >>> >>> to
> >>> >>> take other use cases into account if their requirements overlap
> >>> more
> >>> >>> or less with video.I am a bit worried that your description is too
> >>> >>> far away from that use case. What do others think?
> >>> >>
> >>> >>
> >>> >> Although the examples given by Rub√f¬©n are quite specific to still
> >>> >> images, it seems to me that a similar kind of concern exist for
> >>> >> video: video can be digitalized from analog media, captured by
> >>> >> digital devices or generated; they can be altered in several ways
> >>> >> (re-encoding, subtitling, montage...).
> >>> >
> >>> > Good point. I think an implementation of this is to separate actors
> >>> > or roles like creator, changer and consumer. This is what the
> >>> metadata
> >>> > working group deliverable does, see section 2 of
> >>> > http://www.metadataworkinggroup.com/pdf/mwg_guidance.pdf
> >>> > However what you are mentioning and what Ruben describes sounds to 
> >>> > me
> >>> > rather like a requirement than a use case, that is the requirement 
> >>> > to
> >>> > take such roles into account for relating various metadata
> >>> > vocabularies. What do you think?
> >>> >
> >>> > Felix
> >>> >
> >>> >
> >>> >
> >>>
> >>>
> >>>
> >>
> >>
> >>
> >>
> >>
> >
> >
>
>
>
>
Received on Tuesday, 4 November 2008 17:41:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 4 November 2008 17:41:36 GMT