ADM-201 dump PMP dumps pdf SSCP exam materials CBAP exam sample questions

用深度不透明贴图实时描绘头发 – 译学馆
未登陆,请登陆后再发表信息
最新评论 (0)
播放视频

用深度不透明贴图实时描绘头发

Real-Time Hair Rendering With Deep Opacity Maps | Two Minute Papers #171

亲爱的学霸们 这里是由Károly Zsolnai-Fehér带来的两分钟论文
Dear Fellow Scholars, this is Two Minute Papers with Károly Zsolnai-Fehér.
在之前的视频里 我们已经见过不少关于头发模拟和渲染的视频片段
In earlier episodes, we’ve seen plenty of video footage about hair simulations and rendering.
今天我们来看一个酷炫的新技术
And today we’re going to look at a cool new technique
它能够产生毛发的自投影效果
that produces self-shadowing effects for hair and fur.
在这组对比图片中 你能看得出来这个效果
In this image pair, you can see this drastic difference that shows how prominent this effect
用在头发的外观上有多么巨大的不同
is in the visual appearance of hair.
看到了吧
Just look at that.
多么完美
Beautiful.
但是计算出这种效果的计算量是巨大的
But computing such a thing is extremely costly.
我们有了一片密集的几何图形 例如这成百上千缕发丝
Since we have a dense piece of geometry, for instance, hundreds of thousands of hair strands,
我们必须知道每一缕发丝是怎么挡住其他发丝的
we have to know how each one occludes the other ones.
这会花费超级长的时间来计算
This would take hopelessly long to compute.
至少为了让程序在合理时间内运行
To even get a program that executes in a reasonable amount of time,
我们显然需要进一步简化问题
we clearly need to simplify the problem further.
一个早期技术是利用一些平面来将大量头发分割出层次
An earlier technique takes a few planes that cut the hair volume into layers.
这些平面通常自光源处有规律地向外以一定间隔排列
These planes are typically regularly spaced outward from the light sources
计算一小撮头发要远比计算整个几何图形简单
and it is much easier to work with a handful of these volume segments than with the full geometry.
用越多的平面 我们就得到更多分层
The more planes we use, the more layers we obtain,
预期结果的质量也会更高
and the higher quality results we can expect.
然而 即使我们能够实时运算
However, even if we can do this in real time,
当我们用了大概16层时 就会产生不切实际的图像
we will produce unrealistic images when using around 16 layers.
当然了 我们接下来就要启用更多层平面了
Well of course, we should then crank up the number of layers some more!
如果这样做 比如用128层 我们预期效果会更好
If we do that, for instance by now using 128 layers, we can expect better quality results,
但是我们一张图每秒只能处理2层 这实在太慢了
but we’ll be able to process an image only twice a second, which is far from competitive.
即便如此 最终结果仍然有分层效应
And even then, the final results still contain layering artifacts
和现实差距太远
and are not very close to the ground truth.
肯定有个更好的方式
There has to be a better way to do this.
这项新技术就是深度不透明贴图
And with this new technique called Deep Opacity Maps,
将这些分层的选择更加优化
these layers are chosen more wisely,
这样一来 你只需要用3层就能获得更高品质的结果
and this way, we can achieve higher quality results with only using 3 layers,
并且这种技术运算相当快
and it runs easily in real time.
而且这比先前技术的存储效率更高
It is also more memory efficient than previous techniques.
这个技术的主要想法是如果我们从光源处观察头发
The key idea is that if we look at the hair from the light source’s point of view,
我们就能记下从光源处到几何图像的不同部分的距离有多远
we can record how far away different parts of the geometry are from the light source.
然后 我们根据这些记录的距离 从近到远来创建新的分层
Then, we can create the new layers further and further away according to this shape.
这样 这个分层就不再是二维的了它们适用于我们手头的情景
This way, the layers are not planar anymore, they adapt to the scene that we have at hand
包含远远更有用的遮挡信息
and contain significantly more useful occlusion information.
如你所见 这项新技术把所有之前的技术都比下去了 而且它极其简便
As you can see, this new technique blows all previous methods away and is incredibly simple.
我在Philip Rideout找到了一个应用
I have found an implementation from Philip Rideout,
该应用的链接就在视频简介里
the link to this is available in the video description.
如果你们找到了更多的应用 请务必告诉我我会将你们的发现添加到视频简介中
If you have found more, let me know and I’ll include your findings in the video description
好让诸位好奇心旺盛的学霸看到
for the fellow tinkerers out there.
这篇论文相比之下更为充分你们也一定要看一看哦
The paper is ample in comparisons, make sure to have a look at that too.
有时候 我收到一些信息说"Károly 你为什么总是报道
And sometimes I get some messages saying “Károly, why do you bother covering papers
那些很多年前的论文呢 这根本没道理啊"
from so many years ago, it doesn’t make any sense!”.
这就是两分钟论文的魅力点之一了
And here you can see that part of the excitement of Two Minute Papers is that the next episode
下一期的内容可能是关于任何东西的
can be about absolutely anything.
这一系列基本上都是集中在计算机图画和机器学习的论文上
The series has been mostly focusing on computer graphics and machine learning papers, but
但别忘了 我们也有过一期视频是关于我们是否生活在模拟世界中
don’t forget, that we also have an episodeon whether we’re living in a simulation, or
或者达克效应还有其他的
the Dunning-Kruger effect and so much more.
我已经在视频简介中放了关于这两个的链接 供你观看
I’ve put a link to both of them in the video description for your enjoyment.
报道旧论文的另一个原因是很多人都不知道这些
The other reason for covering older papers is that a lot of people don’t know about them
假如我们能够帮上一点点 来确保
and if we can help just a tiny bit to make sure
这些精妙的成果能够有机会被广泛应用
these incredible works see more widespread adoptions,
我们就已经干得不错了
we’ve done our job well.
感谢你的观看和慷慨支持下期再见
Thanks for watching and for your generous support, and I’ll see you next time!

发表评论

译制信息
视频概述

用来描绘头发在光源下的自投影效果的新技术

听录译者

收集自网络

翻译译者

One静茹

审核员

审核团O

视频来源

https://www.youtube.com/watch?v=7x2UvvD48Fw

相关推荐