Resampling Wheel - Artificial Intelligence for Robotics Udacity English Japanese

Resampling Wheel - Artificial Intelligence for RoboticsYouTube↗ IA↗(long shot)
Education 11.4K views 14 likes 0 dislikes 29 May 2012

Udacity 651K subscribers United States YouTube↗

Click to Play Video

[Narrator] So, here's an idea how to make this more efficient,

これが効率的にするアイデアの1つで

and it turns out empirically it also gives better samples.

よりよいサンプルを実験的に与えてくれることも分かります

Let's represent all our particles and importance weight in a big wheel.

すべての粒子と重要度重みを大きな輪で表しましょう

Each particle occupies a slice that corresponds to its importance weight.

各粒子がその重要度重みに相当する 1つのスペースを占めます

Particles with a bigger weight, like W5, occupy more space.

W5のような重さが大きい粒子はより多くのスペースを占めます

Whereas particles with a smaller weight occupy less space.

一方で重みが小さい粒子はさらに小さなスペースを占めます

Very initially let's guess a particle index uniformly from the set of all indices.

まず最初にすべてのインデックスから均等に粒子のインデックスを推測します

I did note this as a uniform sample at U

1からNの個々の選択肢から

from the discrete set of choices of index 1 all the way to N,

Uの均一なサンプルであることに気づきました

and as a caveat in Python, of course, it goes from 0 to N-1.

Pythonでは0からN－1であることに注意します

So, say we pick W6.

W6を選ぶとします

Then, the trick is--then you're going to construct the function better.

これで関数を上手に構築できるようになるでしょう

Then, I initialize the 0 and to which I add--when I construct these particles--

そして0を初期化して粒子を構築する時に

a uniformly drawn continuous value that sits between 0 and 2 times W max,

この0と2×最大重みWの間に位置する均一に抽出された連続値を加えます

which is the largest of the importance weights in the important set.

重要度の集合において重さが一番大きいものです

W5 is the largest, so we're going to add a random value that might be as large as twice W5.

W5は最も大きいので W5の2倍の乱数を加えます

Suppose the value we added brings us to here.

加えた値はこの位置まで導いてくれます

So, this is the value we actually drew,

これが実際に抽出した値で

measured from the beginning of the sixth particle which shows an initialization.

初期化を示す6番目の粒子から観測されます

I now then iterate the following loop:

そして次のループを繰り返します

if the importance weight of the present particle doesn't suffice

この粒子の重要度重みが

to reach all the way to beta.

βに到達するのに足りないとします

So, if W index isn't as big as beta, then I subtract from beta this very value W index

インデックスWがβほど大きくない場合は βからこのWの値を引いて

and I increment index by 1.

1だけ増加させます

So, what have I done? I've moved index to over here,

これでインデックスをここに動かしました

and I removed this part of beta so the point over here is still the same as before.

βのこの部分を取り除いたので位置は前と同じままです

We now get to the point where beta becomes smaller than W index,

これでβがインデックスWよりも小さくなる位置に到達しました

which is the case in the next situation.

次の状況における場合でインデックス＝７です

Now index=7. Then, index is the index of the particle I pick in my resampling process.

インデックスは再サンプリングステップで選んだ粒子のものです

So, I picked the particle index; I now iterate I add another uniform value to beta.

粒子インデックスを選んだら別の均一な値をβに加えます

Say I add this one. This is the value I add, this is the value beta previously had.

ここでβの前回の値を加えます

The same iteration now will make index flow up

同じ繰り返しによってインデックスが増え

reducing beta by all the slice over here, which is W7,

W7のこの部分すべてにおいてβが減ります

and then jump over here, and particle 1 is picked.

そしてジャンプして粒子1が選ばれるのです

It can easily happen that the uniform value is so small

均一の値がとても小さいので

that the same particle is picked twice, and it's easy to see

同じ粒子が 2度選ばれるということがよく起こります

that each particle is now picked in proportion to the total circumference

各粒子は全外周に比例して選ばれ

it spans in this wheel of particles.

この粒子の輪の中まで及びます

So, this is essentially my implementation for the resampling step.

これが再サンプリングステップの基本的な実装です

So, I want you--if you can--to implement that specific resampler in Python.

もし可能ならPythonで特定の再サンプリングを実装してみてください

Similar videos:

Orientation 2 - Artificial Intelligence for Robotics
Udacity
9.4K views 5 likes 0 dislikes 29 May 2012100% match

Exact Motion - Artificial Intelligence for Robotics
Udacity
15.1K views 12 likes 1 dislikes 19 Feb 2012100% match

Systematic Bias - Artificial Intelligence for Robotics
Udacity
10.9K views 4 likes 0 dislikes 30 May 2012100% match

Parameter Update 2 - Artificial Intelligence for Robotics
Udacity
13.2K views 1 likes 0 dislikes 29 May 2012100% match

Twiddle - Artificial Intelligence for Robotics
Udacity
13.8K views 64 likes 2 dislikes 18 Mar 2012100% match

Evaluate Gaussian - Artificial Intelligence for Robotics
Udacity
15.6K views 4 likes 0 dislikes 29 May 2012100% match

Circular Motion - Artificial Intelligence for Robotics
Udacity
17.2K views 19 likes 2 dislikes 23 Feb 2015100% match

Inexact Move Function - Artificial Intelligence for Robotics
Udacity
17.3K views 10 likes 0 dislikes 19 Feb 2012100% match

Tracking Intro - Artificial Intelligence for Robotics
Udacity
31.9K views 44 likes 0 dislikes 02 Mar 2012100% match

Parameter Optimization - Artificial Intelligence for Robotics
Udacity
11.7K views 8 likes 0 dislikes 18 Mar 2012100% match

Bayes' Rule - Artificial Intelligence for Robotics
Udacity
12.7K views 8 likes 4 dislikes 02 Mar 2012100% match

Another Prediction - Artificial Intelligence for Robotics
Udacity
15.1K views 2 likes 0 dislikes 29 May 2012100% match

Kalman Prediction - Artificial Intelligence for Robotics
Udacity
10.5K views 6 likes 1 dislikes 29 May 2012100% match

Predict Function - Artificial Intelligence for Robotics
Udacity
15.2K views 4 likes 0 dislikes 29 May 2012100% match

Sense Function - Artificial Intelligence for Robotics
Udacity
21.9K views 12 likes 1 dislikes 19 Feb 2012100% match

Maximize Gaussian - Artificial Intelligence for Robotics
Udacity
16.9K views 6 likes 0 dislikes 29 May 2012100% match

Maze - Artificial Intelligence for Robotics
Udacity
9.9K views 4 likes 0 dislikes 29 May 2012100% match

Robot Particles - Artificial Intelligence for Robotics
Udacity
12.1K views 6 likes 0 dislikes 29 May 2012100% match

Kalman Filter Prediction - Artificial Intelligence for Robotics
Udacity
18.2K views 10 likes 7 dislikes 29 May 2012100% match

Coin Flip Quiz - Artificial Intelligence for Robotics
Udacity
21K views 13 likes 0 dislikes 19 Feb 2012100% match

Localization - Artificial Intelligence for Robotics
Udacity
7.4K views 6 likes 2 dislikes 02 Mar 2012100% match

Kalman Filter Design
Udacity
21K views 84 likes 12 dislikes 29 May 2012100% match

Graph SLAM - Artificial Intelligence for Robotics
Udacity
39.1K views 196 likes 9 dislikes 31 May 2012100% match

Planning Solution - Artificial Intelligence for Robotics
Udacity
620 views 0 likes 0 dislikes 16 Mar 2012100% match

Formal Definition of Probability 3 - Artificial Intelligence for Robotics
Udacity
11.5K views 5 likes 0 dislikes 19 Feb 2012100% match

A* in Action - Artificial Intelligence for Robotics
Udacity
21.4K views 236 likes 2 dislikes 30 May 2012100% match

Preferred Gaussian - Artificial Intelligence for Robotics
Udacity
15.7K views 1 likes 0 dislikes 29 May 2012100% match

Variance Comparison - Artificial Intelligence for Robotics
Udacity
16K views 4 likes 2 dislikes 29 May 2012100% match

Normalize Distribution - Artificial Intelligence for Robotics
Udacity
20.5K views 11 likes 4 dislikes 19 Feb 2012100% match

Gradient Descent - Artificial Intelligence for Robotics
Udacity
96.4K views 628 likes 76 dislikes 21 Mar 2012100% match

Settings Privacy About Contact Donate