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
ArtNet - IBM OpenPOWER Cognitive Cup Contest Winning talk
Search
Praveen Sridhar
November 15, 2016
Technology
1
860
ArtNet - IBM OpenPOWER Cognitive Cup Contest Winning talk
ArtNet - IBM OpenPOWER Cognitive Cup Contest Winning talk
Praveen Sridhar
November 15, 2016
Tweet
Share
More Decks by Praveen Sridhar
See All by Praveen Sridhar
Attribute - Python library for Neural Network Interpretability
psbots
0
260
Keep calm and trust your model - On Explainability of Machine Learning Models
psbots
0
140
Talk on Innovation and Entrepreneurship at Rajagiri Business School, Kochi
psbots
0
41
Don't wait to Start-Up Start Exploring already!
psbots
0
130
Firefox OS
psbots
0
170
Firefox OS Indic Keyboard Google Summer of Code Project
psbots
0
470
Other Decks in Technology
See All in Technology
2024/4/26 コンピュータ歴史博物館解説告知
toshi_atsumi
0
210
FrontDoorとWebAppsを組み合わせた際のリダイレクト処理の注意点
kenichirokimura
1
470
Azure犬駆動開発の記録/GlobalAzureFukuoka2024_20240420
nina01
1
190
アクセス制御にまつわる改善 / Improving access control
itkq
0
500
オーナーシップを持つ領域を明確にする
konifar
13
3k
Postman v10リリース後を振り返る / Looking back at Postman v10 after release
yokawasa
1
150
サーバー間 GraphQL と webmock-graphql の話 / server-to-server graphql and webmock-graphql
qsona
2
170
Compose Compiler Metricsを使った実践的なコードレビュー
tomorrowkey
1
210
ChatworkのSRE部って実は 半分くらいPlatform Engineering部かもしれない
saramune
0
150
コードを書く隙間を見つけて生きていく技術/Findy 思考の現在地
fujiwara3
27
5.8k
ワールドカフェI /チューターを改良する / World Café I and Improving the Tutors
ks91
PRO
0
110
現代CSSフレームワークの内部実装とその仕組み
poteboy
8
3.5k
Featured
See All Featured
Designing for humans not robots
tammielis
248
25k
Intergalactic Javascript Robots from Outer Space
tanoku
266
26k
5 minutes of I Can Smell Your CMS
philhawksworth
199
19k
Design and Strategy: How to Deal with People Who Don’t "Get" Design
morganepeng
115
18k
How to train your dragon (web standard)
notwaldorf
72
5.2k
Designing for Performance
lara
601
67k
Building a Scalable Design System with Sketch
lauravandoore
455
32k
It's Worth the Effort
3n
180
27k
In The Pink: A Labor of Love
frogandcode
138
21k
Fireside Chat
paigeccino
20
2.6k
How GitHub Uses GitHub to Build GitHub
holman
468
290k
KATA
mclloyd
14
12k
Transcript
ArtNet The new age Art Connoisseur ;) OpenPower Cognitive Cup
Winning Entry
Developer Challenge Experience Praveen Sridhar @psbots
About me Praveen Sridhar @psbots Machine Learning R&D for Recently
joined as Machine Learning Engineer at
Cognitive Cup Challenge Problem Can computers think “deeply” :P about
art?
What ArtNet Does ArtNet “knows” what a painting is about.
Be it a landscape or still life or a portrait, or any of the different art genres. Examples of genres : Still Life Cityscape Religious
Implementation ArtNet uses • a Convolution Neural Network to figure
out patterns in a painting • and classifies it into different genres
Implementation It uses the Keras Deep Learning library to train
the CNN model
Implementation The OpenPOWER Deep Learning Distribution Frameworks like Theano and
Tensorflow available as Pre-built binaries optimized for GPU acceleration Adding Keras was a breeze, since it runs on top of Theano or Tensorflow
Hyperparameter Optimization The Real Deal HYPER WHAT?
Hyperparameter Optimization The Real Deal Enter the “hyperas” library A
very simple convenience wrapper around hyperopt for fast prototyping with keras models.
Enter the “elephas” library Distributed Deep Learning with Keras &
Spark Thought : SuperVessel Cloud makes it a breeze to spin up multi node spark instances, why not use it as Nitrox! ;)
Going Forward • Predict the artist given a painting •
Help Social Science people in understanding art influences : ✴ which artist influenced whom? ✴ to what extent?
Thanks! Praveen Sridhar @psbots