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

AI根据图像制作3D模型 – 译学馆
未登录,请登录后再发表信息
最新评论 (0)
播放视频

AI根据图像制作3D模型

AI Creates 3D Models From Images | Two Minute Papers #186

亲爱的学霸们 这是由Károly Zsolnai-Fehér带来的两分钟论文
Dear Fellow Scholars, this is Two Minute Papers with Károly Zsolnai-Fehér.
今天 我们要讨论的一项工作 是人类非常擅长的 但是学习算法
Today we’re going to talk about a task that humans are remarkably good at, but learning
却觉得非常麻烦的
algorithms mostly flounder.
那就是用2D彩色图像来创建3D几何结构
And that is creating 3D geometry by looking at a 2D color image.
在视频游戏和动画电影中 这是一个常见的情形
In video games and animation films, this is a scenario that comes up very often – if we
如果游戏需要一件新的武器模型 我们通常会给艺术家提供一张照片 然后他们利用
need a new weapon model in the game, we typically give the artist a photo, who will sit down
一个3D建模程序 花几个小时塑造出 一个与其相似的3D几何结构
with a 3D modeler program, and spends a few hours sculpting a similar 3D geometry.
我接下来就会告诉你 完成这件事并不一定需要我们的肉眼视觉
And I will quickly note that our binocular vision is not entirely necessary to make this happen.
我们可以看一整天2D图像 即使闭上一只眼睛
We can look at 2D images all day long and still have a good idea about the shape of
也依然可以描绘出飞机的形状
an airplane, even with one eye closed.
关于这个问题 我们之前做过一些视频 发现早前的
We had previous episodes on this problem, and the verdict was that that the results
科技成果已经很棒了 只是不够详细
with previous techniques are great, but not very detailed.
数学家喜欢说 这个算法的复杂度呈三次方 或者与三次方成比例
Mathematicians like to say that this algorithm has a cubic complexity or cubic scaling, which
这意味着 即使我们只想将3D模型的分辨率提高一点点 多耗费的时间
means that if we wish to increase the resolution of the 3D model just a tiny bit, we have wait
却不会只是一点点 而是相当长
not a tiny bit longer, but significantly longer.
三次方意味着 即使只是适中的高分辨率 增加的时间耗费也让人无法忍受
And the cubic part means that this tradeoff becomes unbearable even for moderately high resolutions.
本论文提供了一种突破这种限制的技术
This paper offers a technique to break through this limitation.
这种新技术仍然使用学习算法来预测几何结构
This new technique still uses a learning algorithm to predict the geometry,
但是使用分层的方式创建这些3D模型
but it creates these 3D models hierarchically.
意思是 开始时先给出一个大概的接近最终结果的几何结构
This means that it starts out approximating the coarse geometry of the output, and restarts
通过向其添加更多更精致的细节来重复这一过程
the process by adding more and more fine details to it.
这个几何结构在几个步骤后变得越来越细致
The geometry becomes more and more refined over several steps.
只有一个精心设计的算法 才能使这种细致化真正奏效
Now, this refinement doesn’t just work unless we have a carefully designed algorithm around it.
在每一步骤中 利用已创建模型的附加信息来进行细致化
The refinement happens by using additional information in each step from the created model.
也就是说 我们将预测的3D几何看作是许多个小块的集合
Namely, we imagine our predicted 3D geometry as a collection of small blocks, and each
把每个小块分类为 未占据空间 已占据空间以及表面
block is classified as either free space, occupied space, or as a surface.
分类之后 我们可以集中力量将模型表面细致化
After this classification happened, we have the possibility to focus our efforts on refining
使得在算法的执行时间方面有了极大的改善
the surface of the model, leading to a significant improvement in the execution time of the algorithm.
因此 我们得到的3D模型 质量高于以前的技术
As a result, we get 3D models that are of higher quality than the ones offered by previous techniques.
虽然结果还达不到超高分辨率 但它们已经可以捕捉到相当多的表面细节
The outputs are still not super high resolution, but they capture a fair number of surface detail.
钻研是一个过程 每一篇论文都是一块垫脚石
And you know the drill, research is a process, and every paper is a stepping stone.
而这一块垫脚石 能够为3D艺术家们节省
And this is one of those stepping stones that can potentially save many hours of work for
数小时的工作时间
3D artists in the industry.
感谢您的收看和大力支持 我们下期再见!
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

想知道AI是如何将普通图像制作成3D模型吗?快来看看吧!

听录译者

收集自网络

翻译译者

[B]无牙无耳

审核员

审核团O

视频来源

https://www.youtube.com/watch?v=BjwhMDhbqAs

相关推荐