RE: Of interest to the HM group.

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

The VHML work is of enormous interest to some of 
us working on this project.  A few of us come 
from the VRML community and some of our concepts 
are very similar to VHML, dealing with high level 
languages for animation of human figures (avatars) 
and their application to vrml-lit (interactive 
fiction) as well as talking head applications 
(eg, news sites).

Since the VHML work was reviewed in 
phase 0, some feedback and questions for Andrew:

1.  The language is high level, yes?  That is good because 
it is easy to learn and apply.

2.  The language could use HumanML information by 
transformation or an interface.

3.  We were interested in the conformity of the 
VHML models to the H-anim standards given both 
the ISO status of VRML and the presence of VRMLies 
on this list.

4.  On first and not a very deep analysis, it 
seems that the VHML models are convincing but 
also western-centric.  A key problem of inter-cultural 
communications is that gestures such as smile, nod, 
etc. are not universal.  In some cases, a semantic 
mapping might be needed to ensure the gesture is 
communicating the right message.   

We discussed syntax for adding HumanML attribute types to 
languages such as VHML.  Given namespaces, that 
should be straightforward.   We also discussed 
genre-languages that could use languages such as 
VHML to create presentations that followed the 
genre rules (for example, plot points or schedules 
based on necessary events).  We elicited a genre 
description from a local news anchor to create 
an example of such a presentation.

Another project of interest if the EMOTE work so 
it would be good to know if the VHML work has 
reviewed that.  

Len
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: raytrace@smtp.cs.curtin.edu.au
[mailto:raytrace@smtp.cs.curtin.edu.au]
Sent: Friday, September 07, 2001 1:46 AM
To: rkthunga@humanmarkup.org
Cc: raytrace@smtp.cs.curtin.edu.au;
humanmarkup-comment@lists.oasis-open.org
Subject: Of interest to the HM group.




Hi
  >
Joe Norris wrote:
  >Much attention has been drawn to the myriad applications for HumanMar=
  >kup and it's associated specifications and technologies. Considering =
  >the nearly inexhaustible scope for HumanMarkup endeavors I would like=
  >to propose that somehow within the OASIS TC process we develop a sub=
  >committee/process/steering group/workflow management/etc that can ref=
  >ine the process for determining research efforts for various HumanMar=
  >kup applications.

As a way of introduction to this list, I am currently researching an XML
based language for directing a Virtual Human. I am a full partner in a
European Union 5th Framework project called InterFace that is concerned
with Human Computer Interface issues and one aspect of this is the
production of computer generated Talking Heads. Our group has developed and
implemented VHML - a Virtual Human Markup Language. This language marks up
text that the Talking Head would read and adds tags such as <happy>,
<look_left> <smile> etc so that the voice and the head displays these
emotions/gestures. You may find examples of audio produced this way as
well as movies of the Talking Head at www.vhml.org
The emotion and gesture tags are quite convincing.
For those interested, we also have journal articles, conference papers and
upcoming book chapters which talk about the language.

An initial draft of the VHML document can also be found at that site
although this is really an in-house version which is currently undergoing
modification for verification/validation of the language. It uses the W3C
format and has added to the W3C document describing the Speech Synthesis
Markup Languauge (i.e it blatantly uses that document since the Text to
Speech sub-language of VHML uses SSML). The next version of the document
should be available before the Sept 17th meeting.

I have monitored this list for a while and I am not certain if what we do
is appropriate to what you talk about but we are committed to evolving our
language to cater for the types of things that a Human would be able to
do. The HM group seems to be more interested in descriptive rather than
directive markup but the two do come from the same basic knowledge.

I am happy to provide my expertise in this very practical area of
Human Markup for any Technical Committee. If the powers that be are
interested in what we do and what we have learnt and can use that
expertise, then please contact me and I will be happy to assist.

My next 5th Framework meeting in Europe is at the end of September and I
intend to report on what the Human Markup group is doing and how we can
fit in with their efforts.