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
The fun Deep Learning
Search
Takuya Matsuyama
October 30, 2015
Technology
0
2.8k
The fun Deep Learning
Takuya Matsuyama
October 30, 2015
Tweet
Share
More Decks by Takuya Matsuyama
See All by Takuya Matsuyama
ネイティブモジュールの作り方 @ React Native Meetup #9 in Japan
craftzdog
6
1.3k
How to Create Native Modules @ React Native Japan Meetup #9
craftzdog
1
860
Introducing Inkdrop for Mobile Built with React Native
craftzdog
1
2.2k
Caffeでお手軽本格ディープラーニングiOSアプリ
craftzdog
1
1.6k
Other Decks in Technology
See All in Technology
AIは変更差分からユニットテスト_結合テスト_システムテストでテストすべきことが出せるのか?
mineo_matsuya
5
2.7k
テストを実行してSorbetのsigを書こう!
sansantech
PRO
1
130
[OCI Technical Deep Dive] OCIで生成AIを活用するためのソリューション解説(2025年8月5日開催)
oracle4engineer
PRO
0
130
Android Studio の 新しいAI機能を試してみよう / Try out the new AI features in Android Studio
yanzm
0
110
歴代のWeb Speed Hackathonの出題から考えるデグレしないパフォーマンス改善
shuta13
6
550
僕たちが「開発しやすさ」を求め 模索し続けたアーキテクチャ #アーキテクチャ勉強会_findy
bengo4com
0
2.6k
Amazon Qで2Dゲームを作成してみた
siromi
0
180
Engineering Failure-Resilient Systems
infraplumber0
0
130
はじめての転職講座/The Guide of First Career Change
kwappa
5
4.5k
Mackerel in さくらのクラウド
cubicdaiya
1
340
AWS DDoS攻撃防御の最前線
ryutakondo
1
180
生成AI活用のROI、どう測る? DMM.com 開発責任者から学ぶ「AI効果検証のノウハウ」 / ROI of AI
i35_267
4
130
Featured
See All Featured
Exploring the Power of Turbo Streams & Action Cable | RailsConf2023
kevinliebholz
34
6k
Optimising Largest Contentful Paint
csswizardry
37
3.4k
[Rails World 2023 - Day 1 Closing Keynote] - The Magic of Rails
eileencodes
36
2.5k
Making Projects Easy
brettharned
117
6.3k
Distributed Sagas: A Protocol for Coordinating Microservices
caitiem20
333
22k
Mobile First: as difficult as doing things right
swwweet
223
9.9k
Gamification - CAS2011
davidbonilla
81
5.4k
Fashionably flexible responsive web design (full day workshop)
malarkey
407
66k
Designing for Performance
lara
610
69k
Being A Developer After 40
akosma
90
590k
Designing Experiences People Love
moore
142
24k
KATA
mclloyd
32
14k
Transcript
The fun Deep Learning TAKUYA (@noradaiko) Design Japan
Takuya Matsuyama দࢁ • Twitter: @noradaiko • Freelance since
4 years ago ‣ ex. Yahoo! JAPAN • iOS app, web app developer • Love drawing & playing drums 2
3
walknote music discovery app 4
VTFST 5
Men-kiki ໙ར͖ 6
There are.. a lot of ramen shops in Japan!! 7
Difficult to read & input especially for foreigners 8 Ұའ
ఱԼҰ Ѩར ࢁ಄Ր େউݢ ݹλϯϝϯ ೋ Ұ෩ಊ
can offer ramen shops based on a PHOTO 9
10 +*30TUZMF *OQVU
11 :PLPIBNBGBNJMZTUZMF *OQVU
M e n - k i k i DEMO 12
13 Men-kiki uses Deep Learning technology ʴ ʹ
BCSBODIPGTPDBMMFE"* XIJDIIBT IJHIBDDVSBDZJOJNBHFSFDPHOJUJPO 14 %FFQMFBSOJOHJT
• Inspired by nervous systems of the animal brain •
It can learn features of objects from huge amount of data automatically 15
16 SBNFOJNBHFT -FBSO bXIBUSBNFOJT`
M e n - k i k i ໙ར͖ 17
* English version will be available soon
Deep learning can be also used for.. ART 18
A Neural Algorithm of Artistic Style by Leon A. Gatys,
Alexander S. Ecker, and Matthias Bethge 19 + style
A Neural Algorithm of Artistic Style by Leon A. Gatys,
Alexander S. Ecker, and Matthias Bethge 20 + style
Deep learning can clone the painters’ sense 21
Interesting! 22 + style 1JDBTTP .ZQIPUP 3FTVMU
How about VIDEO? It must be beautiful…
Let’s do it 1. Record videos which will match artistic
style 2. Convert all its frames into artistic style 24
None
None
Wrap up • Deep learning is so fun • Let’s
make great stuff with it!! • Follow me(@noradaiko) on Twitter 27