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

使用相位神经网络来实时控制人物 – 译学馆
未登录,请登录后再发表信息
最新评论 (0)
播放视频

使用相位神经网络来实时控制人物

Real-Time Character Control With Phase-Functioned Neural Networks | Two Minute Papers

亲爱的学霸同学们 这里是Károly Zsolnai-Fehér带来的两分钟论文
Dear Fellow Scholars, this is Two Minute Papers with Károly Zsolnai-Fehér.
在这篇论文里 让我们一起来探索如何实时控制虚拟人物
In this piece of work, we seek to control digital characters in real-time.
它是这样实现的: 我们先指定一条目标轨迹 然后这种算法
It happens the following way: we specify a target trajectory, and the algorithm has to
必须综合一系列的动作来跟随这条轨迹运动
synthesize a series of motions that follows that path.
为了使这些动作尽可能的真实 典型的做法是
To make these motions as realistic as possible, this is typically accomplished by unleashing
将一种学习算法置于一个巨大的含有无数多的动作信息的数据库中
a learning algorithm on a large database that contains a ton of motion information.
以前的技术对于这些数据库没有很好的认识 而且通常
Previous techniques did not have a good understanding of these databases and they often synthesized
将对应不同种类运动的动作综合在一起
motions from pieces that corresponded to different kinds of movements.
这样就导致了输出的动作僵硬、不自然
This lack of understanding results in stiff, unnatural output motion.
做个直观的类比 这就像用从不同报纸上剪下的一个个字母
Intuitively, it is a bit like putting together a sentence from a set of letters that were
拼成一句话
cut out one by one from different newspaper articles.
它虽然是一个完整的句子 但是却缺少那种
It is a fully formed sentence, but it lacks the smoothness
工整对齐的文本所带来的平滑流畅
and the flow of a properly aligned piece of text.
这篇论文介绍的是一种基于神经网络的技术 它将相位函数引入到学习过程中
This is a neural network based technique that introduces a phase function to the learning process.
这种相位函数增加了对于所给动作的时间信息的学习
This phase function augments the learning with the timing information of a given motion.
有了这种相位函数 神经网络就不仅仅认识到我们在学习周期性动作
With this phase function, the neural network recognizes that we are not only learning periodic motions,
还知道这些动作该何时开始和结束
but it knows when these motions start and when they end.
这项技术消耗内存很少 可以实时运行
The final technique takes very little memory, runs in real time,
它非常成功地完成了在各种地形上的流畅的行走、跑步、
and it accomplishes smooth walking, running, jumping and climbing motions
跳跃、攀爬以及其它很多动作
and so much more over a variety of terrains with flying colors.
在之前的节目中 我们讨论过运用另一种技术通过一个低等和高等控制器
In a previous episode, we have discussed a different technique that accomplished something
来完成类似的事情
similar with a low and high level controller.
这项技术最主要的卖点之一是 在穿越不同地形时 只需运用一个神经网络
One of the major selling points of this technique is that this one offers a unified solution
就能得到一个统一的解
for terrain traversal with using only one neural network.
这样就使得我们非常有潜力做出相当大的计算机游戏和实时动画
This has the potential to make it really big on computer games and real-time animation.
我能够见证这些并且成为未来的一部分 实在是让人惊叹
It is absolutely amazing to witness this and be a part of the future.
请务必看一下这篇论文 因为其中还包含了地形的拟合步骤细节
Make sure to have a look at the paper, which also contains the details of a terrain fitting step
以及如何使这个学习算法将各种障碍物也考虑进来
to make this learning algorithm capable of taking into consideration a variety of obstacles.
在此 我想感谢Claudio Pannacci将那么多集节目
I would also like to thank Claudio Pannacci for his amazing work
都翻译成了意大利文
in translating so many of these episodes to Italian.
这使得两分钟论文能够被全球更多的人看到
This makes Two Minute Papers accessible for more people around the globe,
越多的人能看到 我就越高兴
and the more people we can reach, the happier I am.
感谢收看和支持 下次再见
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

本集介绍了一种基于神经网络的技术,它将相位函数引入到学习算法中,从而可以实时控制虚拟人物在各种地形上流畅的运动。

听录译者

收集自网络

翻译译者

豆子

审核员

知易行难

视频来源

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

相关推荐