Telepresence continued

[Jan Forbrich]

Hi all,

I would like to get back to the telepresence topic (in its widest
definition) that we briefly discussed... a while ago (this is a test on
how many old emails you keep!). If you are surprised about this email,
it just means that you are subscribed to the low-energy astrophysics
group. I hope this email finds the group reasonably well since it hasn't
exactly been overly active. I have kept the thoughts below in a
classical text file for an even longer while, and I thought I should
post them before text files become obsolete, having some renewed
interest in experimentation related to telepresence/-conferencing. Also,
I'm still somewhat baffled to see more and more refined teleconferencing
techniques in business, but not really(?) in science. It would be great
to hear your comments. This email has two topics,
telecast journal clubs and telecast talks.

1) Telecast journal clubs

Phil, you mentioned Sarah Bridle who tried to put together a journal
club via skype. I think this is a really great idea, and it is a bit sad
to hear that this apparently did not work out. (Is Sarah here on this
list? Hello!) I think there are two quite different types of journal
club, and one of them could possibly gain from being done via skype or
something similar.

There is the general journal club that many smaller astronomy groups
seem to have. This is of interest to learn more about what is happening
outside your immediate (sub-)subfield, even though I'm not really sure
how useful that really is since normally the discussions either by
necessity have to remain superficial or they leave out significant
fractions of attendees. Then, there are topical journal clubs, like on
star formation, for example. These at least sometimes generate very good
discussions, and more easily so than after a talk, partially because "it
is easier to criticize than to be correct". This type of journal club
depends on there being a critical mass of people working in at least
roughly the same field. In many cases, and in particular at smaller
institutes, that is not the case.

I thus think that a topical journal club might actually benefit from
some sort of internet interaction, even though I'm not sure a mass skype
is the best way to do it. If one would bring together, for example, a
star formation journal club, spanning institutions, that could, at least
potentially, be better than most local topical journal clubs could be.
I'm tempted to try that out, starting with a couple of friends, just to
find out what works and what does not. Maybe the most efficient way
would be to link several smaller topical journal clubs into a larger one
via the Internet? (You might form a local topical journal club on
something with just one other person, for example, if you knew that the
discussion is generated and sustained by web interaction. This would
also mean that it is not a number of people each on their own in front
of skype.) This would be different from simply broadcasting a journal
club on the web where you may never know whether the author concerned is
eavesdropping on what you say about him/her...

Anyway - anyone interested in experimenting in any type of telecast
journal club, please let me know! (Or, even better, please let me/us
know if you know about one! Google seems to suggest that such things exist!)

2) Telecast talks

It was said that meetings are not primarily for the exchange of
information, but rather for schmoozing, which is probably true. This
means that they may not be the best venue to start thinking about
implementing any kind of telepresence, even if that depends on the type
of meeting. Arguably, colloquia are more geared toward exchange of
information (you'd invite someone to present some new results... and
then maybe schmooze afterwards). I have been thinking for a while that
it would be cool to have a website with an up-to-date listing of live
webcasts of colloquia. More and more places these days webcast at least
some of their colloquia, even if for various different reasons. (Some,
like NRAO, have various sites that are meant to get together, others not
enough space to get everyone into one room, etc.). If there was an easy
place to go to and look at a listing of live webcasts, I think I would
occasionally watch one.

At the same time, there could be links to archived talks which may still
be of similar interest if they are posted soon after the actual talk
(and you may not want to follow the live webcast at 4am in your local
timezone). There may be less need for a central website here, but, for
example, I would appreciate learning about good recorded talks in other
fields of astronomy (which is currently only happening randomly). The
best place to list such archived talks may actually be the ADS! A good
recent example of how to post-process conference talks before publishing
them I think is this: http://origins.physics.mcmaster.ca/oi_planets/

Obviously, this would be a bit like watching TV and an interactive part
after a telecast talk would be a lot more difficult to establish, at
least if you don't know the speaker's cell phone number. On the positive
side, watching TV means that you can switch off a talk that turns out to
be less interesting than expected (or zap to a different one...). But
just in order to get used to the idea and establish telecast talks more,
I think this would be a good starting point.

Looking forward to hearing your thoughts.

Enuff. Servus from Austria,
Jan

[B@LBL]

Jan, thank you very much for your email. 

I think I have written about this before (as I am a victim of several projects on different continents, and therefore we are all victims of my carbon emission during the travel ) so, I will just briefly recap my issues with telepresence tools and offer a possible improvement path. 

Problems:  Telepresence tools are very unsatisfactory thus far, especially because they don't include the laser pointer.

 1. In the old-style telecon tools, one could never ever read the viewgraphs, making the experience very low value.  There was good interaction with the speaker, either explicitly included in the main video or in picture-in-picture, however.

2. In new-style gotomeeting.com type presentations, you see the powerpoint screen, you hear speaker, and you can call in on the 800 number for true audio interaction.   The problem here is that the presenter cannot indicate or emphasize points on the screen with a pointer.  This is a serious limitation, and for most of the presentation, there is no reason to not simply see an after-the-fact video or powerpoint.   In fact, questions have very limited utility.  Unless the presenter can go back five slides and use the laser pointer to ask if some item on some plot was what the questioner wanted, and so forth, it's much less valuable then being there in person. 

My conclusion is that the most important problem with telepresence tools is lack of a laserpointer for viewers. 

Solving the Problem, improving telepresence to reduce carbon emission

-------------------------------------------------------------------------

Obviously with money and time and special equipment, this can be solved.  However, after years of the modern generation of gotomeeting-like tools, it has not been solved. 

I just saw an interesting talk today by Prof. Kazuo Hemmi, Director of Information and Media Studies, University of Nagasaki, Japan.   He has simple programs that use image differencing for very, very fast response on very ordinary hardware that are very robust.  He has applied them to catching motion and overlaying video; his demo is to make stars appear  as people on camera move their fingertips around in the air in front of them, and the stars robustly follow their fingertips.   His camera is a $10 webcam, his computer is  a nondescript laptop.  Some of his code runs on java or other languages that run through your browser.   Now imagine on the speakers computer Prof. Hemmi's software gets the webcam input from the speakers computer and the video output of powerpoint.  As long as the  speaker gets the feedback of where his pointing finger goes on the powerpoint frame, mission accomplished:  The remote viewer would see the combination of the  powerpoint with an overlay of a pointing symbol moving with the real speaker's finger (could be a blinking star, or whatever you like), so the speaker can literally point out whatever he want to emphasize in their presentation.  (Note 1)

Skype Overlay for personal interactions:  I also have a lab project on another continent.  I can have my student, on skype,  point the camera at the apparatus, but it would be much, much more useful if *I* could point somewhere on the image for my student to see.  Again, this could be done as an overlay using my finger motion as input, and transmitting back out skype somehow. 

Let's do something if we can do anything

---------------------------------------------------------

FTP was around forever, but it took physicists to make computer communication much more useful by inventing the WWW.  We shouldn't wait around for telepresence companies to make the required improvements.  We should do what we can to lead them. 

I wrote an email to prof. Hemmi.  Perhaps some representative from DOE or NASA could communicate some of this to gotomeeting.com?  Or to skype? Perhaps this would give some impetus for improvements? 

What do you all think? 

-Bruce

(1) In some kind of fancier version, the webcam on the live powerpoint screen could be registered appropriately to the pointpoint virtual screen, the Hemmi program could easily follow a real laser pointer and project it back on the transmitted powerpoint for the viewer to see.  

UC Berkeley Space Sciences Laboratory

Bruce_...@lbl.gov

510-486-5489

fax - please send scans by email

--
This message was sent to the "low-energy-astrophysics" Google group.
To post to this group, send email to LEa...@googlegroups.com
Visit the group at http://groups.google.com/group/LEastro?hl=tk
Visit our wiki at http://low-energy-astro.physics.ucsb.edu

To unsubscribe, send email to LEastro+u...@googlegroups.com

[Benjamin Weiner]

Hi all,

A collaboration I work with uses Readytalk which essentially
allows one person to show their desktop - it appears in a
Flash container in a web browser - and has a separate pane
that lists the meeting attendees, allows one to send a virtual
"raise hand" signal, and so on. I believe that along with the
desktop it shows the cursor, so that can be used as a pointer.
Since it's showing a desktop, you can use it to show anything -
premade slides, but also text editor windows, notes someone
is typing, code, etc. It is possible, I think, to give a remote
user control of the pointer, but this needs to be used with care.

I don't really think that the pointer is the most critical issue; it is
an annoyance though.
I think the largest problem with telepresence is that it removes all
of the subtle cues that indicate how the audience is reacting,
when someone wants to speak, and so on. The virtual "raise hand"
is an attempt to put some of those cues back in, but it's much
harder for the presenter to keep track of. You almost need a
presenter and an emcee who pays attention to both the lead
presenter and the audience. It is hard to make virtual presence
as good as real presence because we've all got decades of
experience interpreting subtle cues from real conversation partners.

I expect we are all familiar with the telecon phenomenon that there
are a small number of people who are more willing to start talking
and talk over the others who begin sentences more slowly, and
that these people dominate the interaction. Also that when one
does get to start talking, one has to keep talking until reaching the
end of an idea. If you pause for breath you may not get the microphone
back. It's like that damned shell in "Lord of the Flies."

Again this is an issue that is difficult to solve by applying yet more
technology, because we're trying to substitute technology for manners.

Ben

--
Benjamin Weiner
Assistant Astronomer, Steward Observatory
b...@as.arizona.edu
http://mingus.as.arizona.edu/~bjw/

[Phil Marshall]

I use vyew.com to present slides remotely, and that has a pointer that
can be seen by the remote audience (in fact, anyone viewing can point
at things, each cursor is labelled with its user's name!). It's a nice
system - I used it to give an invited talk at SnowPAC in March. I
should really blog that experience...

If you want to view my slides, check out
http://vyew.com/743311/SnowPAC2012 . Unfortunately the pointer is only
there when I am in that window! Take my word for it though. BTW vyew
was significantly faster than sharing the keynote window over skype.

According to the SnowPAC audience, the talk went well - from my host
laptop's webcam I could see the audience and their raised hands, but
not their facial expressions well. UCL have an excellent videocon room
with very good cameras and a big screen for the speaker. You can see
*their* face better than you would be able to in real life... My
impression is that with a bit of effort it should be able to put
together a pretty effective remote seminar setup.

Re journal clubs, there was (is?) a long running weak lensing journal
club in Europe, attended by people in Edinburgh, London, Paris, Bonn
and other places too. I think they use(d) EVO, which is nice, but in
my experience not as able to cope with multiple video feeds as well as
skype premium or google+. I like the G+ hangouts best actually - they
offer screen and window sharing. Brendon Brewer and I are currently
working on a paper together via this system (he's in Auckland now).

Cheers

Phil

Dr. Phil Marshall

Department of Physics (Astrophysics)
University of Oxford,
Denys Wilkinson Building, Room 532E (BIPAC)
Keble Road, Phone: +44 1865 273345
Oxford, OX1 3RH http://www.slac.stanford.edu/~pjm

[Jan Forbrich]

Hi Phil and others,

thanks for your replies. Maybe telepresence actually was the wrong title
for my email and it should instead be more general 'web interactivity'
or something. It is interesting to learn about new tools as they become
available and how they are being used. It is certainly a matter of
taste, to some degree. Maybe it would be worth collecting links to these
tools on our wiki? (A while ago, a friend of mine pointed me to another
interesting tool... bubbl.us for collaborative brainstorming and
'mind-mapping'. A simple tool, but a good idea... in the web world, the
tool is already very old, though.)

I think that one can look at this from two different perspectives. On
the one hand, we can look at this from a viewpoint where we try to find
appropriate tools for activities like meetings which traditionally were
done without technology. In that case, it is difficult to exactly
emulate that experience, even though it is interesting to see how close
to at least some key purposes the use of technology can get (
telepresencerobots.com :-D ). On the other hand, we can look at this
from a viewpoint where we try to adapt activities to the available
tools, focusing on key purposes that can be realized with existing
technology. That's what I meant when saying that, for example, topical
journal clubs may actually benefit from web interactivity.

It may be interesting to see specific examples of successful
implementations. For example, the NASA Herschel Science Center organizes
webinars on technical data reduction issues with live interaction using
Webex. I've participated in a few of them and thought they were very
successful. Again - this is a type of meeting where I think web
interactivity can be used very effectively (if it is not a week-long
meeting, maybe): Even though it's always nice to meet people, the main
purpose of these technical workshops is information exchange, and that
worked very well with Webex (with presentations plus data analysis
sessions with desktop sharing).

Bruce, I like your point of how it took scientists to invent the world
wide web (if it wasn't Al Gore :-) ). It might indeed be an interesting
exercise to try to define an 'ideal' web interactivity tool, from a
scientist's perspective. Maybe there are too many contradictory ideas
around, but maybe not. One could do this for a list of specific
purposes. In fact, I've witnessed a similar process that took place in a
foundation that promotes intercultural communication via the internet (
www.meltonfoundation.org ). The whole thing started in the early
nineties when even using email was highly innovative. When it became
clear that much more web interaction was possible beyond email, we tried
to define an ideal 'communication system' for our purposes. After the
project definition, a company was selected to do the implementation. The
process was not easy and in the end resulted in something like a
Facebook clone, but one of the many interesting lessons was that most of
the required software was available on an open-source basis, so it was
actually technically easier than initially expected. My personal
impression is that it may be too early for such a science web
interactivity tool, since at least in astronomy (how about other areas
of science?) the whole idea is not really established. Also, it would be
good to try existing tools first and find out about their strengths and
weaknesses. Maybe we indeed need a collaborative brainstorming tool to
begin with...!

Servus from Vienna,
Jan