vrWhales

main research page with papers here

announcements

Random Thoughts on Exhibit

• Could cover surface in ice, with a small break somewhere for belugas to breathe. This would solve some of the problems that happen when belugas encounter surface.

• If we use beluga cam could run it on slow motion.

• Steve suggested having multiple (invisible) targets that get ‘turned on’ at various times to move belugas to different areas.

• Interested in scenario of positioning visitors as beluga researchers watching wild belugas.

• Emphasis on sound, could we use some of sounds from Venus Project? Would require Toggling sound on/off or adjusting volume through scripting.

Wild Beluga Resources

Good site on wild beluga research (especially on Lancaster Sound pods): www.whales-online.net Aquarium’s Arctic Curriculum grades 4-7: http://www.vanaqua.org/education/documents/Part3AABACKGROUNDINFO.pdf

Aquarium side project (not beluga related but focus on BC whales)http://www.wildwhales.org/stewardship/index.html

Notes on V7 with scripting

• Belugas are no longer attracted to target, with the exception of the calf, so belugas are never in scene and get stuck going to surface.

• I just realized that in previous versions, calf was attracted to mother, but mother does not seem to be attracted to calf. Would be good to add this in, as well as some aggressive behaviour fom mother when male tries to attack calf. Also if multiple females present, which female is mother? Is this coded in .ini?

• Calf sex is set as child (not sure if this matters or not), but may be a constraint in future versions.

• All belugas set with initial fear/aggression = 0. Sp far haven’t been able to change fear/aggression levels with scripting.

• Need help function to list all scripting commands

• Would be good to add any behaviours in progress as additional output for name ?.

• Best that screen not freeze while typing in commands in later versions, but is useful for debugging now.

• Found help feature by pressing ctrl+h, describes slow motion and other commands.

Notes from Aquarium Visit on Thursday June 14, 2007

Purpose: Testing of vBeluga on large screen in Environmental News area, followed by meeting with Jeff and Catherine.

• We were able to show belugas on large screen, but were unsure what resolution they were being shown at. Picture was not as crisp and bright as computer.

• Two new screens have been installed on left and right side walls. On left wall is a large rectangular screen (projector currently has wrong lens), on right is a long inverted rectangular screen (using temp projector).

• Jeff suggested possibility of running multiple screens with multiple computers.

• Sound in Environmental News can either be directional from single screen or general from all speakers.

• Discussed two location possibilities for VBeluga exhibit: Environmental News and Arctic Area. If in Environmental News, Jeff suggested it should be run as an interpreter show, if in Arctic it should be a standalone show (that could be run without interpreters or volunteers or else I would be the one running it). Potential third possibility would be isolated testing in classrooms with summer camps.

• If we did an interpreter run show, we would have to contact Don Enright, Manager of Interpreters to work out logistics. Catherine also suggested that they could assign a volunteer to work on this project.

• Theme of aquarium until October is Pollution. Then after that it will be Arctic. Both of which fit perfectly with the exhibit. Focusing on pollution for this exhibit would be ideal, environmental or sound pollution.

Notes from Aquarium Visit on Friday May 25, 2007

• Steve, Caitlin and Veronica met with Jeff, Catherine and the content group to present VBeluga and Face projects.
• Content group has only been complete for a month, so they are very busy getting up to speed and creating content for environmental news area
• Jeff suggested that belugas could be shown in environmental news area (on large screen) and potentially a show could be developed around it for the large theatre
• Both content group and Eric agreed that it was ok if vbeluga wasn’t perfect, they were interested in trying new things on the floor, and it would provide valuable feedback.
• Jeff suggested it would be good to demonstrate behaviours from ethogram with info viz (or interpreter) to explain what behaviours visitors are seeing, so they can go back to live belugas and see them in a new way.
• Suggested that educational messages had to be very short 30 seconds to 2 minutes
• Aquarium interested in placing more emphasis on vBeluga as a tool for research
• Jeff mentioned Venus Project, pointing out that they have mp3’s of underwater sounds that are interesting.

After meeting did a tour of DEC. Observed that many of the exhibits showed a disconnect between the animals on display and the technology/activity associated with it. We thought of several ways that connection could be made more explicit. Also, environmental news area is completely empty right now, and doesn’t draw visitors in. It also has the constraint that it needs to be used for events in the evening, so it cannot have any permanent furniture, and any furniture in there would need to be stored during events.

Notes from meeting at UBC on Friday May 18, 2007

• Viewed demo of belugas on smartscreen
• Noticed that V. 6 is quite good, implements breathing, and has less collision detection problems and improvements to chest area over V.5
• Noted that there are still some issues with collision particularly with beluga tails (likely due to lack of sensors on lower half of belugas). Would it be possible to add sensors to lower half?
• The surface still presents problems to belugas, they don’t have any behaviour associated with surface and get pushed back by the surface which can lead to some unnatural looking behaviour.
• Surface waves are also too large, a calmer surface would make it easier to see belugas when they surface. Also there is not enough transparency at the surface.

• Mentioned that it would be nice to have beluga cam (i.e. view scene from the viewpoint of a beluga, like in butterfly fish demo) as it would be engaging to visitors

• Water effects (caustics) can be distracting, we should try to tone them down a bit in the .ini files

• Would be interesting to take a narrative approach to the interface. One possibility would be to position viewer as beluga researchers, who are watching wild belugas from a submarine. They can move the camera view, and move the ball as an attractor to the belugas, future iterations might allow food to be released. This leads to interesting info viz possibilities, such as having a display that shows when a ship is approaching, or possibly showing animal names & characteristics. Could provide buttons that allow food to be dispersed or toys to be moved.

• Caitlin to write project proposal

Proposed Timeline for Caitlin’s Summer RA

• lit review
• concept. proto
• low level testing
• dev robust working proto
• user testing in aq
• paper

Scripting for VBeluga

V. 6

Currently the available interaction is:

• Camera movement (up, down, left, right, backwards, forwards)
• Movement of target forward and backwards
• Right click on beluga for context info (currently their name)
• Control-shift right click - scene properties dialog (background, ambient light, fog, light position, wireframe)
• Control-right click - beluga properties dialog (yaw, pitch,fluke controller, left/right fin controller, head controller, linear/angular drag, mouth control, dynamics, autonomous, forces, sensors
• Control-left click on target moves it forward, left click moves it back

.ini files # Naming convention for .ini files: beluga123.ini, 1 = # females, 2 = # calves, 3 = # males

• beluga type (0-calf, 1-female, 2-male)
• beluga starting position (x,y,z)
• beluga starting direction (y)
• beluga scale (female 1.0, calf smaller, male bigger)
• beluga name

environment variables

• camera position
• background colour
• fogcolor
• ambient light
• light
• shadow
• foglevel
• space
• shadows
• caustics
• caustic
• navigation bar
• target starting position
• surface characteristics

Current Behaviours

All:

• breathing

Male

• aggressive

o jaw-clapping

        o chasing calf
  * more independent than females, less attracted to target

Female

• interaction with ball (looks like playing with ball)
• mother/calf interactions?
• more grouping/attraction to target

Steve’s suggestion for scripting:

Imagine a command line interface:

- list

    tovuc - young, male
    soso - old, female, mother
    jojo - middle aged, male

- jojo : aggression ?

     jojo aggression is 50

- jojo: aggression = 100

     jojo's aggression is 100

- jojo list_all

      name   age        sex     aggression   fear  happy  breathing

oooo oooo

      jojo      young     male   100              50    10       .5
         .5       .3

- tovac: attachmother soso

     tovac mother is: soso

- tovac: fear = 100

     tovac's fear is 100

- tovac: follow soso .4

     tovac set to follow soso with closeness level of .5

- wait 30 - tovac: unfollow

     tovac: unfollows

Minimum scripting requirements:

If we just control beluga state (as above):

• Need to add additional beluga properties (young, male or old, female, mom, etc) to .ini files
• Need to know more about beluga states. What state variables are there and what are their ranges?

    Examples of states: attachment level, aggression, hunger, attraction, fear, happiness, playfulness
  * Would we need to script behaviours? Or are behaviours so closely tied to state that we could change one state variable, and that would trigger behaviours associated with that state?

Other Possible extensions to vBeluga - not related to scripting

• Add other behaviours: playing with target (already partially there), increased mom/calf interaction (possibly feeding-calf nuzzles up to mom, she slows down or stops swimming so calf can drink), group swim, feeding behaviour (could change target into food or ice), bubbleplay
• vocalizations for adult beluga (playful and aggressive sounds available)
• transparency mode is really evocative, would be nice to have it again in future versions

Issues:

• Collision detection problems, belugas sometime swim right through each other.
• When you turn off the dynamics and autonomy of a whale they seem to cease to exist to the other whales who swim past and through that whale like they aren’t there. Though they clearly have some indication that the whale is there because I once turned off the dynamics and autonomy of a mother, and her calf came and started swimming in circles right through her.
• Sometimes after you turn off dynamics/autonomy then turn them back on, the beluga seems to get stuck. They move, but they no longer go anywhere. This is probably not high priority because only we use this feature.
• Problems when the whales surface and seem to get pushed back, so it can sometimes take several tries before they are able to get away from the surface.