[x3d-public] [x3d] Spec Comment by dougsanden on 19774-2: HAnim Motion Capture

[Don Brutzman]

Thank you Doug.  Sounds perfect.  Let's share screens.  You're on the agenda!  8)

On 5/27/2020 8:25 AM, GPU Group wrote:
> Don,
> Discuss Friday - yes sure maybe 5 minutes worth or so. I'm not sure how to present. If you can push some of my links, that's good enough.
> - Keep in mind I still have Motion implemented as spec'd.
> - So these other things are 'extra' and don't wreck what's already specified - you can still have Motion as spec'd.
> - except what might be different for v4 based on this: the abstract type(s) Motion inherits:
> Motion: MotionPlay, MotionData
> -Doug
> 
> On Wed, May 27, 2020 at 8:14 AM Don Brutzman <brutzman at nps.edu <mailto:brutzman at nps.edu>> wrote:
> 
>     [cc: x3d-public]
> 
>     Doug, thanks for this work.  It is interesting to consider that we might be able to map the HAnim specification's Motion object more effectively into X3D4 than the baseline HAnimMotion approach.
> 
>     [1] HAnim 2 Specification, Parts 1 and 2
>     https://www.web3d.org/documents/specifications/19774/V2.0/index.html
> 
>     [2] Humanoid Animation (HAnim): Motion Data Animation, Concepts
>     https://www.web3d.org/documents/specifications/19774-2/V2.0/MotionDataAnimation/MotionDataAnimation.html
>     https://www.web3d.org/documents/specifications/19774-2/V2.0/MotionDataAnimation/Concepts.html
> 
>     [3] Humanoid Animation (HAnim) 19774-2 part 2: clause 6 HAnim motion data animation using Motion objects
>     https://www.web3d.org/documents/specifications/19774-2/V2.0/MotionDataAnimation/MotionNodes.html
> 
>     [4] Extensible 3D (X3D) 4.0 Tooltips: HAnimMotion
>     https://www.web3d.org/x3d/content/X3dTooltips.html#HAnimMotion
> 
>     [5] Web3D Mantis: view issues
>     https://www.web3d.org/member-only/mantis/view_all_bug_page.php
> 
>     TODO: Mantis issue needed.
> 
>     Wondering, might you present this issue during Friday morning's X3D Working Group teleconference?  Discussion would help, interested attendees welcome.  If you like we can record video too - whatever works.  Thanks for considering these possibilities.
> 
>     v/r Don
> 
>     On 5/27/2020 5:40 AM, Spec Feedback wrote:
>      > -- Submitter indicates that this comment may be public: *Yes* --
>      >
>      > Comment on 19774-2: HAnim Motion Capture - V1.0
>      > 4.4.4 mapping and Motion
>      >
>      >
>      > -----------------
>      > In previous comments I proposed having a split of Motion into MotionPlay and
>      > MotionData/MotionDataFile nodes. With MotionDataFile and downloaing .bvh
>      > mocap motion capture files off the internet, there are 2 issues
>      > 1. mapping of joint names
>      > 2. accumulating rotations over skipped/ignored joints
>      > This comment is For issue#1, joint name mapping. In a previous comment I
>      > enumerated ways to map:
>      > 1. compiled in common name mappings
>      > 2. geometrical mapping
>      > 3. explicit name mapping
>      > This comment is for method #3 explicit name mapping.
>      > Currently: in our HH HAnimHumanoid we can name each joint whatever we want.
>      > And in the Motion.joints field, we need to use the same naming scheme. The
>      > Motion node and HH are in the same scene file and a single scene author would
>      > author both nodes.
>      > Proposed: for the split MotionData/MotionDataFile it will be handy to have
>      > some way to re-use motions downloaded off the internet with multiple x3d
>      > scenes without having to with name lookup permutations for each combination
>      > of HH and MotionDataFile.
>      > One way to add consistency / universality is to map HH names to a consistent
>      > benchmark naming scheme, and same with downloaded bvh.
>      > And one way to do that is with a field or node for mapping to LOA1, LOA2,
>      > LOA3 standard joint names.
>      > - each HH would have this mapping field/node to get from scene-authored joint
>      > names to standard LOA names
>      > - and each Motion, MotionData, MotionDataFile would also have a mapping from
>      > its joint names to standard LOA names.
>      > Then the code would do 2 lookups when matching joints to motions:
>      > 1. HH name to standard LOA name
>      > 2. standard LOA name to MotionData joint name
>      > This 2=step process would allow different HH to be paired with different
>      > MotionDataFile/MotionData without the scene author needing to redo naming for
>      > each permutation - they would do mappings for each once.
>      > -----------------
>      >
>      > Submitted on Wednesday, 2020,  May 27 - 6:40am
>      > by dougsanden (dougsanden )
>      > IP: 75.159.18.239
>      >
>      > See: https://www.web3d.org/node/1694/submission/4018
>      > _______________________________________________
>      > x3d mailing list
>      > x3d at web3d.org <mailto:x3d at web3d.org>
>      > http://web3d.org/mailman/listinfo/x3d_web3d.org
> 
>     all the best, Don
>     -- 
>     Don Brutzman  Naval Postgraduate School, Code USW/Br brutzman at nps.edu <mailto:brutzman at nps.edu>
>     Watkins 270,  MOVES Institute, Monterey CA 93943-5000 USA   +1.831.656.2149
>     X3D graphics, virtual worlds, navy robotics http://faculty.nps.edu/brutzman
> 

all the best, Don
-- 
Don Brutzman  Naval Postgraduate School, Code USW/Br       brutzman at nps.edu
Watkins 270,  MOVES Institute, Monterey CA 93943-5000 USA   +1.831.656.2149
X3D graphics, virtual worlds, navy robotics http://faculty.nps.edu/brutzman