This seems pretty crude. Particularly in that it ignores visceral fat entirely. Seems like a huge sacrifice in the accuracy of the output. The only acceptable application I could think of for a model with this level of detail would be for EA's "The Sims" game.
Meta discussion is one of the core strengths of HN comments. If the GP felt that this was the most important takeaway from the article, then it's good that s/he didn't force others to read 9 dense pages of prose in order to gain exposure to the idea. And really, the title mentioning bodybuilding implies that this is a practical model. The GP is pointing out that it probably isn't.
The research presented is useful for far more than models for the "EA
Sims game" as was disparagingly claimed. The title is also fitting since
the methods presented are applicable to quickly building practical
models of bodybuilding. Though the modeling of visceral/organ fat is
specifically excluded for further research, the fat modeling methods
presented are both effective and useful for model generation. Their use
of a quasi-static solver ("Projective Dynamics") on both their muscle
and fat tissue growth modeling to get "near-interactive" run times is
particularly impressive. The single paragraph "conclusion" section of
the paper shows the paper holds interesting, useful, and innovative
research:
>"To our knowledge, our work is the first to simulate physics-based
growth processes of human tissues in computer graphics. We believe
that our system will be instrumental in reducing the often prohibitive
costs of human body modeling and will find applications even beyond the
traditional realms of computer graphics, such as film, games, and visual
effects."
Of course, a lot of people won't have the time to read a long academic
paper, so comment quantity and quality often suffers on submissions of
academic works, but there's a problem here that you seem to be missing.
The comment by 'sigmar' is actually just another example of the long
standing "mean and/or dumb" comment problem on HN [1]. The comment is
both factually wrong and unnecessarily derogatory. That comment should
have been at least down-voted, and possibly even flagged, rather than
defended as some imaginary "core strength" of HN.
The phrase "mean and/or dumb" is from PG, but it's a old and common
problem affecting most discussion forums. Manipulative people regularly
make failed attempts to look smart in public discussions by being
excessively negative, and sadly, it tends to work in so much as lots of
other people are fooled into up-voting them. Anyone can be a totally
uninformed and acrimonious critic, and unfortunately, most human beings
are complete suckers for controversy, so the "mean and/or dumb but
massively up-voted" comment problem tends to be self-reinforcing. The
only good news is, both the bad commenters and the bad up-voters are
quietly handled in code [2].
When you see the "mean and/or dumb" comments, or even the excessively
negative comments, you should down-vote them since they are against the
site guidelines [3, 4]. When a comment is particularly egregious, click
the "X hours ago" link on the comment, and then the 'flag' link (if it's
available). Like the down-vote privilege, the flag privilege also has a
karma threshold to prevent abuse.
When you realize how the excessively negative comments are really just
trying to seek attention, you understand they are just another form of
trolling. In this case, 'sigmar' is doing simple drive-by-trolling; say
something as offensive as possible to get attention, and refuse to back
it up with references when questioned about it. Yep, you got trolled.
But I hope you'll try to look for this sort of negative nonsense in the
future, and down-vote/flag it appropriately.
There is trolling happening here, however, it's not coming from sigmar who offered a terse criticism and moved on. In that light, this will be my final post on this thread.
I put in the effort to disagree with you politely and provide a detailed
explanation with references. Though it seems you've written off my
response as trolling, I hope you remember and reconsider it. You'll be
better off in the long run if you do.
>In this case, 'sigmar' is doing simple drive-by-trolling; say something as offensive as possible to get attention, and refuse to back it up with references when questioned about it.
How is the game reference "offensive"? Their own article offered up "games" as an application of the research (in the section you quoted).
Furthermore, no one asked me a single question about my comment.
You must have a ton of time on your hands to go around trolling like this.
Wow, that's some great research there. We are actually building a scale for consumers that will allow for body scans in 3D. We hope then to also later integrate morphing algorithms so that you can see how your body would transform if you were to gain extra muscles and fat. Potentially big data could also help.
As for the visceral fat, it's indeed there, but an MRI could probably help off-set such, if you wanted to have a high level of accuracy.
For most I believe outside fat will be most important.
10 comments
[ 3.0 ms ] story [ 29.5 ms ] threadThe research presented is useful for far more than models for the "EA Sims game" as was disparagingly claimed. The title is also fitting since the methods presented are applicable to quickly building practical models of bodybuilding. Though the modeling of visceral/organ fat is specifically excluded for further research, the fat modeling methods presented are both effective and useful for model generation. Their use of a quasi-static solver ("Projective Dynamics") on both their muscle and fat tissue growth modeling to get "near-interactive" run times is particularly impressive. The single paragraph "conclusion" section of the paper shows the paper holds interesting, useful, and innovative research:
>"To our knowledge, our work is the first to simulate physics-based growth processes of human tissues in computer graphics. We believe that our system will be instrumental in reducing the often prohibitive costs of human body modeling and will find applications even beyond the traditional realms of computer graphics, such as film, games, and visual effects."
Of course, a lot of people won't have the time to read a long academic paper, so comment quantity and quality often suffers on submissions of academic works, but there's a problem here that you seem to be missing.
The comment by 'sigmar' is actually just another example of the long standing "mean and/or dumb" comment problem on HN [1]. The comment is both factually wrong and unnecessarily derogatory. That comment should have been at least down-voted, and possibly even flagged, rather than defended as some imaginary "core strength" of HN.
The phrase "mean and/or dumb" is from PG, but it's a old and common problem affecting most discussion forums. Manipulative people regularly make failed attempts to look smart in public discussions by being excessively negative, and sadly, it tends to work in so much as lots of other people are fooled into up-voting them. Anyone can be a totally uninformed and acrimonious critic, and unfortunately, most human beings are complete suckers for controversy, so the "mean and/or dumb but massively up-voted" comment problem tends to be self-reinforcing. The only good news is, both the bad commenters and the bad up-voters are quietly handled in code [2].
When you see the "mean and/or dumb" comments, or even the excessively negative comments, you should down-vote them since they are against the site guidelines [3, 4]. When a comment is particularly egregious, click the "X hours ago" link on the comment, and then the 'flag' link (if it's available). Like the down-vote privilege, the flag privilege also has a karma threshold to prevent abuse.
When you realize how the excessively negative comments are really just trying to seek attention, you understand they are just another form of trolling. In this case, 'sigmar' is doing simple drive-by-trolling; say something as offensive as possible to get attention, and refuse to back it up with references when questioned about it. Yep, you got trolled. But I hope you'll try to look for this sort of negative nonsense in the future, and down-vote/flag it appropriately.
[1] https://news.ycombinator.com/item?id=2403696
[2] https://news.ycombinator.com/newsfaq.html
[3] https://news.ycombinator.com/newsguidelines.html
[4] https://news.ycombinator.com/item?id=9317916
How is the game reference "offensive"? Their own article offered up "games" as an application of the research (in the section you quoted).
Furthermore, no one asked me a single question about my comment.
You must have a ton of time on your hands to go around trolling like this.
As for the visceral fat, it's indeed there, but an MRI could probably help off-set such, if you wanted to have a high level of accuracy.
For most I believe outside fat will be most important.