Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
三次元形状とディープラーニング
Search
Tatsuya Yatagawa
December 21, 2020
Technology
0
290
三次元形状とディープラーニング
三次元形状処理とディープラーニングの初歩についてまとめたスライドです。
Tatsuya Yatagawa
December 21, 2020
Tweet
Share
More Decks by Tatsuya Yatagawa
See All by Tatsuya Yatagawa
Event-based Camera Simulation using Monte Carlo Path Tracing with Adaptive Denoising (ICIP 2023)
tatsy
0
530
VC'23 チュートリアル講演 『三次元メッシュと深層学習』
tatsy
0
210
SIGGRAPH 2022: Neural Geometry Processing
tatsy
0
240
Real-time Shading with Free-form Planar Area Lights using Linearly Transformed Cosines (JCGT & I3D 2022)
tatsy
0
980
SIGGRAPH Asia 2021: Real-Time Rendering
tatsy
0
120
SIGGRAPH 2021: Denoising and Guiding
tatsy
0
94
SIGGRAPH 2021: Shape Generation
tatsy
0
170
SIGGRAPH 2021: Meshing
tatsy
0
110
12. Keyboard & Mouse
tatsy
0
210
Other Decks in Technology
See All in Technology
30万人の同時アクセスに耐えたい!新サービスの盤石なリリースを支える負荷試験 / SRE Kaigi 2026
genda
4
1.4k
SREが向き合う大規模リアーキテクチャ 〜信頼性とアジリティの両立〜
zepprix
0
470
Greatest Disaster Hits in Web Performance
guaca
0
280
Red Hat OpenStack Services on OpenShift
tamemiya
0
130
AI駆動開発を事業のコアに置く
tasukuonizawa
1
360
【Ubie】AIを活用した広告アセット「爆速」生成事例 | AI_Ops_Community_Vol.2
yoshiki_0316
1
120
日本の85%が使う公共SaaSは、どう育ったのか
taketakekaho
1
240
Oracle Cloud Observability and Management Platform - OCI 運用監視サービス概要 -
oracle4engineer
PRO
2
14k
コンテナセキュリティの最新事情 ~ 2026年版 ~
kyohmizu
5
1k
Ruby版 JSXのRuxが気になる
sansantech
PRO
0
160
制約が導く迷わない設計 〜 信頼性と運用性を両立するマイナンバー管理システムの実践 〜
bwkw
3
1k
What happened to RubyGems and what can we learn?
mikemcquaid
0
310
Featured
See All Featured
Measuring Dark Social's Impact On Conversion and Attribution
stephenakadiri
1
130
Introduction to Domain-Driven Design and Collaborative software design
baasie
1
590
Lightning talk: Run Django tests with GitHub Actions
sabderemane
0
120
Building a Scalable Design System with Sketch
lauravandoore
463
34k
How to Align SEO within the Product Triangle To Get Buy-In & Support - #RIMC
aleyda
1
1.4k
A better future with KSS
kneath
240
18k
Unsuck your backbone
ammeep
671
58k
End of SEO as We Know It (SMX Advanced Version)
ipullrank
3
3.9k
Avoiding the “Bad Training, Faster” Trap in the Age of AI
tmiket
0
79
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
508
140k
実際に使うSQLの書き方 徹底解説 / pgcon21j-tutorial
soudai
PRO
196
71k
Why Mistakes Are the Best Teachers: Turning Failure into a Pathway for Growth
auna
0
54
Transcript
None
• 機械学習 ニューラルネット 深層学習
• • • θ • = ; = 2 2
+ 1 + 0 ෩ = arg min =0 −1 ( , ; )
• • • 0 = 0 + 0 1 =
1 + 1 2 = 2 + 2 = 0 0 + 1 1 + 2 2 + = 0 0 + 1 1 + 2 2 +(0 0 + 1 1 + 2 2 + )
• • • 0 1 2 0 = ReLU(0 +
0 ) 1 = ReLU(1 + 1 ) 2 = ReLU(2 + 2 )
= 32 = 128 = 512 • •
• • = 512 = 1 = = 4
•
•
•
•
•
•
• • • • • • • • • •
•
(c) Velodyne (c) Andrew Tallon (c) Microsoft (c) Sony
• , , = 0
′ = − ′ = + ′
• ☺ ☺ ☺ ☺ ,
, = 0
☺ ☺ ☺ ☺ , ,
= 0
• • • • •
• • • •
•
•
• ⚫ ⚫ ⚫
• • •
• • •
• •
• • •
• • • • • • •
☺ ☺ ☺ ☺ , ,
= 0
• • • (c) Velodyne
• • • (0 , 0 , 0 ) (1
, 1 , 1 ) ( , , ) (−1 , −1 , −1 ) (+1 , +1 , +1 ) (0 , 0 , 0 ) (1 , 1 , 1 ) (−1 , −1 , −1 ) (+1 , +1 , +1 ) ( , , )
• • •
None
⚫
⚫
• •
• • •
•
•
•
• • • • • • •
☺ ☺ ☺ ☺ , ,
= 0
• • •
• • • •
• • ∗ = ℱ−1(ℱ ⋅ ℱ ) ′ =
ℱ−1(Θ ∘ ℱ )
• = 0 1 1 0 1 0 1 0
1 1 0 1 0 0 1 0 = − − 1 2− 1 2 = :
• • • = T
• () ′ = ℱ−1(Θ ∘ ℱ ) ′ =
ℱ = () ℱ−1 = ()
• • • +1 +2 +1 = ReLU ⋅
• • •
• • •
• •
• • •
• • •
• •
• • • • • •
☺ ☺ ☺ ☺ , ,
= 0
• , , =
• • •
• • • •
• •
• • •
• • •
• • • •
•
• • • •
• • •
• • • •
• • •
• • • •
• • • • • • • •
• • • • • • • • •