SlideShare une entreprise Scribd logo
1  sur  38
Visual Classification
without examples
Classificeren van beeld zonder voorbeelden
Thomas Mensink
VOGIN-IP-LEZING 2015
• What is an axolotl?
• Some examples
Preview
VOGIN-IP 20152
VOGIN-IP 20153
We can classify based on labeled examples
(supervised learning)
Preview
• What is an aye-aye?
• Textual description:
– Is nocturnal
– Lives in trees
– Has large eyes
– Has long middle fingers
VOGIN-IP 20154
VOGIN-IP 20155
We can classify based on a textual description
(and some prior knowledge)
VOGIN-IP 20156
Can a computer do the same?
(yes, that is what this talk is about)
Agenda
• Supervised Visual Classification
• Attribute-Based Classification
• Co-occurrence Based Classification
VOGIN-IP 20157
Computer Vision – in the news
VOGIN-IP 20158
Visual Recognition
9
Cityscap
eOutdoor
…
tree
Buildin
g
Lamp
People
John
Dam
Slide credit: Jan van Gemert
VOGIN-IP 2015
Supervised Classification
• Obtain annotated examples
• Find a representation
• Train a generic classifier
VOGIN-IP 201510
Remarks:
- New class: retrain on new examples
- How to obtain training examples?
- How to represent images?
Visual Classification: Two trends
Datasets
• 2005: motorbikes,
bicycles, people, cars
• Since 2010: ImageNet
15K
Representations
• 2005: Manual derived
features/encodings
• 2012: Trained end-to-
end, Deep Neural Nets
VOGIN-IP 201511
VOGIN-IP 201512
1000 classes, 5 guesses per image
Current state-of-the-art: 6.7% error
Estimating human performance
VOGIN-IP 201513
Andrej Karpathy: “I realized that I needed to go through the painfully
long training process myself”Test set 1500 images
GoogleNet: 6.8% error
Karpathy: 5.1% error
VOGIN-IP 201514
Visual Classification:
near-human performance
when ample train data is available
Attribute-based classification
VOGIN-IP 201515
1. Define vocabulary
2. Train visual classifiers
3. Class to attribute mapping
4. Infer class
What are good attributes?
• Good attributes
– are task and category dependent;
– class discriminative, but not class specific;
– interpretable by humans; and
– detectable by computers
VOGIN-IP 201516
Quiz: What are good attributes?
• is grey?
• is made of atoms?
• lives in Amsterdam?
• is sunny?
• eat fish?
• has a SIFT descriptor with empty bin 3?
• has 4 wheels?
VOGIN-IP 201517
How many attributes?
• In theory k binary attributes can represent
– 2k classes
• In practice for c classes we need
– Many attributes
VOGIN-IP 201518
Animals with Attributes
VOGIN-IP 201519
Animals with Attributes - Vocabular
VOGIN-IP 201520
Class to attribute mapping
VOGIN-IP 201521
Attribute Based Prediction
1. Learn attribute classifiers
from related classes
2. Train and Test set are
disjoint
3. Infer attributes from new
test image
4. Use attribute-to-class
mapping to predict class
VOGIN-IP 201522
Animals with Attributes (results)
VOGIN-IP 201523
Disadvantages
• Unnatural distinction between
– Attributes to be detected
– Classes of interest
• Inherently multi-class zero-shot prediction
VOGIN-IP 201524
Classification based on co-occurrences
I’m looking for a label, which I have not seen
before. However, this picture contains also:
1. Indoor
2. Living room
3. Table
4. Chair
VOGIN-IP 201525
VOGIN-IP 201526
We can classify based on context
COSTA: Design
VOGIN-IP 201527
COSTA: Design
VOGIN-IP 201528
• Many visual concepts can be described as an open
set of concept-to-concept relations
• Describe image semantics with co-occurrences
• Exploit natural bias in natural images
Exploit natural bias in natural images
VOGIN-IP 201529
Sink"is"u
visual"spa
a"stove,"
(2)
onal to the
ments weestimate lab
truthlabellingof our i
occurrencestatisticsc
pora,e.g.,Wordnet or
suchasYahoo, Google
3
Lreg =
i
kwi −
k
akw
=
X
i
X
d
wid − a>
whereindex i and k both run over th
sii = 0. The vector vid contains
weightedweight vectorsvidk = sik w
Notethat thelossisformulatedove
over train images. Moreover, Eq. (9)
=
i d
wid −
whereindex i and k both run ov
sii = 0. The vector vid conta
weightedweight vectorsvidk = s
Notethat thelossisformulated
over train images. Moreover, Eq
obtainedinclosed-formusingrid
weobservethat regularization is
formance, thedimensionality of a
(1)
etween the
s paper, we
atistics be-
erent simi-
two labels.
informativeclueabout thev
isalsoshowninanimagere
In addition to the positi
c++
ij , wealsousetheother p
thepresenceof label i with
senceof label i withthepres
of bothlabels, denotedby c+
i
eachof thesedefinitionsof
larity measuresdefinedabov
Using the positive and
weight vector w of anunkn
COSTA: Classifier
• Goal: Estimate classifier for unseen label
• Knowledge base:
– k trained classifiers
– Co-occurrences
• Zero-shot classifier:
VOGIN-IP 201530
Co-Occurrence Statistics
• Ground-truth data (proof-of-concept)
• Web search engines
• Flickr Tags
• Language resources
• Visual annotated data (eg Microsoft COCO)
VOGIN-IP 201531
Example: Beach Holiday
VOGIN-IP 201532
Concept Normalized Co-Oc Weight
Sea 0.1810
Water 0.0992
Summer 0.0548
LandscapeNature 0.0435
SunsetSunrise 0.0383
Sports 0.0367
Travel 0.0347
Ship 0.0346
Sunny 0.0319
Big Group 0.0282
Example: Beach Holidays
VOGIN-IP 201533
Results per concept
VOGIN-IP 201534
Co-occurrences from the Web
VOGIN-IP 201535
Conclusions
• Supervised visual classification performs well
when ample train data is available
• Classification without examples:
– Define some set of base classifiers
– Transfer new class to space of these classifiers
– Two examples: attributes and co-occurrences
VOGIN-IP 201536
Thanks to:
• The organizers
• Christoph Lampert for slides and inspiration
• Authors of the cited papers
• Colleagues and supervisors (UvA: Amir, Cees, Jan, Spencer &
Stratis, PhD: Cordelia, Florent, Gabriela, Jakob)
VOGIN-IP 201537
Literature
• Frome, Corrado, Shlens, Bengio, Dean, Ranzato,and Mikolov,
“DeViSE: A Deep Visual-Semantic Embedding Model”, NIPS 2013
• Habibian, Mensink, and Snoek, “VideoStory: A New Multimedia
Embedding for Few-Example Recognition and Translation of Events”,
ACM MM 2014
• Lampert, Nickish, and Harmeling, “Attribute-Based Classification for
Zero-Shot Learning of Object Categories”, TPAMI 2013
• Li, Gavves, Mensink, and Snoek, “Attributes Make Sense on
Segmented Objects”, ECCV 2014
• Mensink, Gavves, and Snoek, “COSTA: Co-Occurrence Statistics for
Zero-Shot Classification”, CVPR 2014
• Norouzi, Mikolov, Bengio, Singer, Shlens, Frome, Corrado, and Dean,
“Zero-Shot Learning by Convex Combination of Semantic
Embeddings”, ICLR 2014
VOGIN-IP 201538

Contenu connexe

En vedette

Is intelligence informatie?
Is intelligence informatie?Is intelligence informatie?
Is intelligence informatie?voginip
 
Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.
Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.
Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.Eric Sieverts
 
Onderzoeksdata in beeld / In Search 4 Data
Onderzoeksdata in beeld / In Search 4 DataOnderzoeksdata in beeld / In Search 4 Data
Onderzoeksdata in beeld / In Search 4 DataMarina Noordegraaf
 
Iedereen factchecker; tools en technieken
Iedereen factchecker; tools en techniekenIedereen factchecker; tools en technieken
Iedereen factchecker; tools en techniekenvoginip
 
Social media tools
Social media toolsSocial media tools
Social media toolsvoginip
 
Een nieuwe rol voor IP-ers
Een nieuwe rol voor IP-ersEen nieuwe rol voor IP-ers
Een nieuwe rol voor IP-ersvoginip
 
Inzet van kennisportals tussen organisatie en klant
Inzet van kennisportals tussen organisatie en klantInzet van kennisportals tussen organisatie en klant
Inzet van kennisportals tussen organisatie en klantvoginip
 
De factcheckparadox
De factcheckparadox De factcheckparadox
De factcheckparadox voginip
 
Brave new search world
Brave new search worldBrave new search world
Brave new search worldvoginip
 
Een beter internet voor kinderen
Een beter internet voor kinderenEen beter internet voor kinderen
Een beter internet voor kinderenvoginip
 
Use of Twitter and social media
Use of Twitter and social mediaUse of Twitter and social media
Use of Twitter and social mediavoginip
 
Smartlogic, Semaphore and Semantically Enhanced Search – For “Discovery”
Smartlogic, Semaphore and Semantically Enhanced Search –  For “Discovery”Smartlogic, Semaphore and Semantically Enhanced Search –  For “Discovery”
Smartlogic, Semaphore and Semantically Enhanced Search – For “Discovery”voginip
 
Business case social media bij Gemeenten
Business case social media bij Gemeenten Business case social media bij Gemeenten
Business case social media bij Gemeenten voginip
 
Van bibliometrics naar altmetrics
Van bibliometrics naar altmetricsVan bibliometrics naar altmetrics
Van bibliometrics naar altmetricsWouter Gerritsma
 
Newsreader vogin-ip-26-mar-2015
Newsreader vogin-ip-26-mar-2015Newsreader vogin-ip-26-mar-2015
Newsreader vogin-ip-26-mar-2015Piek Vossen
 
Transforming Web Search: From Keyword to Knowledge Graph
Transforming Web Search: From Keyword to Knowledge GraphTransforming Web Search: From Keyword to Knowledge Graph
Transforming Web Search: From Keyword to Knowledge Graphvoginip
 
Workshop Informatie vindbaar met metadata en taxonomieën
Workshop Informatie vindbaar met metadata en taxonomieënWorkshop Informatie vindbaar met metadata en taxonomieën
Workshop Informatie vindbaar met metadata en taxonomieënJoyce van Aalten
 
Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015
Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015
Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015Joyce van Aalten
 
Meten is weten vogin ip workshop 2016 joyce van aalten invenier
Meten is weten vogin ip workshop 2016 joyce van aalten invenierMeten is weten vogin ip workshop 2016 joyce van aalten invenier
Meten is weten vogin ip workshop 2016 joyce van aalten invenierJoyce van Aalten
 
Is a mobile phone more dangerous than an AK47?
Is a mobile phone more dangerous than an AK47?Is a mobile phone more dangerous than an AK47?
Is a mobile phone more dangerous than an AK47?voginip
 

En vedette (20)

Is intelligence informatie?
Is intelligence informatie?Is intelligence informatie?
Is intelligence informatie?
 
Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.
Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.
Semantisch zoeken - over knowledge graph, semantisch web, rdf enz.
 
Onderzoeksdata in beeld / In Search 4 Data
Onderzoeksdata in beeld / In Search 4 DataOnderzoeksdata in beeld / In Search 4 Data
Onderzoeksdata in beeld / In Search 4 Data
 
Iedereen factchecker; tools en technieken
Iedereen factchecker; tools en techniekenIedereen factchecker; tools en technieken
Iedereen factchecker; tools en technieken
 
Social media tools
Social media toolsSocial media tools
Social media tools
 
Een nieuwe rol voor IP-ers
Een nieuwe rol voor IP-ersEen nieuwe rol voor IP-ers
Een nieuwe rol voor IP-ers
 
Inzet van kennisportals tussen organisatie en klant
Inzet van kennisportals tussen organisatie en klantInzet van kennisportals tussen organisatie en klant
Inzet van kennisportals tussen organisatie en klant
 
De factcheckparadox
De factcheckparadox De factcheckparadox
De factcheckparadox
 
Brave new search world
Brave new search worldBrave new search world
Brave new search world
 
Een beter internet voor kinderen
Een beter internet voor kinderenEen beter internet voor kinderen
Een beter internet voor kinderen
 
Use of Twitter and social media
Use of Twitter and social mediaUse of Twitter and social media
Use of Twitter and social media
 
Smartlogic, Semaphore and Semantically Enhanced Search – For “Discovery”
Smartlogic, Semaphore and Semantically Enhanced Search –  For “Discovery”Smartlogic, Semaphore and Semantically Enhanced Search –  For “Discovery”
Smartlogic, Semaphore and Semantically Enhanced Search – For “Discovery”
 
Business case social media bij Gemeenten
Business case social media bij Gemeenten Business case social media bij Gemeenten
Business case social media bij Gemeenten
 
Van bibliometrics naar altmetrics
Van bibliometrics naar altmetricsVan bibliometrics naar altmetrics
Van bibliometrics naar altmetrics
 
Newsreader vogin-ip-26-mar-2015
Newsreader vogin-ip-26-mar-2015Newsreader vogin-ip-26-mar-2015
Newsreader vogin-ip-26-mar-2015
 
Transforming Web Search: From Keyword to Knowledge Graph
Transforming Web Search: From Keyword to Knowledge GraphTransforming Web Search: From Keyword to Knowledge Graph
Transforming Web Search: From Keyword to Knowledge Graph
 
Workshop Informatie vindbaar met metadata en taxonomieën
Workshop Informatie vindbaar met metadata en taxonomieënWorkshop Informatie vindbaar met metadata en taxonomieën
Workshop Informatie vindbaar met metadata en taxonomieën
 
Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015
Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015
Taxonomieën maken en toepassen in SharePoint 2013 VOGIN-IP lezing 2015
 
Meten is weten vogin ip workshop 2016 joyce van aalten invenier
Meten is weten vogin ip workshop 2016 joyce van aalten invenierMeten is weten vogin ip workshop 2016 joyce van aalten invenier
Meten is weten vogin ip workshop 2016 joyce van aalten invenier
 
Is a mobile phone more dangerous than an AK47?
Is a mobile phone more dangerous than an AK47?Is a mobile phone more dangerous than an AK47?
Is a mobile phone more dangerous than an AK47?
 

Similaire à Voginip lezing 2015: Classificeren zonder voorbeelden

SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用
SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用
SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用SSII
 
MediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video Archives
MediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video ArchivesMediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video Archives
MediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video Archivesmultimediaeval
 
R-Recommenders Matrix Factorization - RUG PH Meetup
R-Recommenders Matrix Factorization - RUG PH MeetupR-Recommenders Matrix Factorization - RUG PH Meetup
R-Recommenders Matrix Factorization - RUG PH MeetupThe Real Dyl
 
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN BarcelonaDeep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN BarcelonaUniversitat Politècnica de Catalunya
 
PredictionIO - Building Applications That Predict User Behavior Through Big D...
PredictionIO - Building Applications That Predict User Behavior Through Big D...PredictionIO - Building Applications That Predict User Behavior Through Big D...
PredictionIO - Building Applications That Predict User Behavior Through Big D...predictionio
 
AI for PM.pptx
AI for PM.pptxAI for PM.pptx
AI for PM.pptxNatan Katz
 
Scikit-Learn: Machine Learning in Python
Scikit-Learn: Machine Learning in PythonScikit-Learn: Machine Learning in Python
Scikit-Learn: Machine Learning in PythonMicrosoft
 
Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...
Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...
Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...YutaSuzuki27
 
7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...
7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...
7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...Skolkovo Robotics Center
 
Csw2016 d antoine_automatic_exploitgeneration
Csw2016 d antoine_automatic_exploitgenerationCsw2016 d antoine_automatic_exploitgeneration
Csw2016 d antoine_automatic_exploitgenerationCanSecWest
 
Detecting Food and Activities in Lifelogging Images
Detecting Food and Activities in Lifelogging ImagesDetecting Food and Activities in Lifelogging Images
Detecting Food and Activities in Lifelogging ImagesRami Albatal
 
Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3
Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3
Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3Dr. Aparna Varde
 
Monitoring at Cloud Scale
Monitoring at Cloud ScaleMonitoring at Cloud Scale
Monitoring at Cloud ScaleJulien Pivotto
 
A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015
A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015
A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015COBWEB Project
 
EDUPUB Implementation Demo Showcase - Reference SW using Readium JS
EDUPUB Implementation Demo Showcase - Reference SW using Readium JSEDUPUB Implementation Demo Showcase - Reference SW using Readium JS
EDUPUB Implementation Demo Showcase - Reference SW using Readium JSOpen Cyber University of Korea
 

Similaire à Voginip lezing 2015: Classificeren zonder voorbeelden (20)

SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用
SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用
SSII2021 [OS2-03] 自己教師あり学習における対照学習の基礎と応用
 
MediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video Archives
MediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video ArchivesMediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video Archives
MediaEval 2015 - SAVA at MediaEval 2015: Search and Anchoring in Video Archives
 
R-Recommenders Matrix Factorization - RUG PH Meetup
R-Recommenders Matrix Factorization - RUG PH MeetupR-Recommenders Matrix Factorization - RUG PH Meetup
R-Recommenders Matrix Factorization - RUG PH Meetup
 
Visual Object Analysis using Regions and Local Features
Visual Object Analysis using Regions and Local FeaturesVisual Object Analysis using Regions and Local Features
Visual Object Analysis using Regions and Local Features
 
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN BarcelonaDeep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
 
PredictionIO - Building Applications That Predict User Behavior Through Big D...
PredictionIO - Building Applications That Predict User Behavior Through Big D...PredictionIO - Building Applications That Predict User Behavior Through Big D...
PredictionIO - Building Applications That Predict User Behavior Through Big D...
 
ECAI-20_KGE_tutorial.pdf
ECAI-20_KGE_tutorial.pdfECAI-20_KGE_tutorial.pdf
ECAI-20_KGE_tutorial.pdf
 
AI for PM.pptx
AI for PM.pptxAI for PM.pptx
AI for PM.pptx
 
Scikit-Learn: Machine Learning in Python
Scikit-Learn: Machine Learning in PythonScikit-Learn: Machine Learning in Python
Scikit-Learn: Machine Learning in Python
 
Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...
Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...
Transfer Learning Model for Image Segmentation by Integrating U-NetPlusPlus a...
 
7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...
7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...
7. Григорьев Илья Презентация питч-сессии Интеллектуальная Робототехника SUV2...
 
Video + Language: Where Does Domain Knowledge Fit in?
Video + Language: Where Does Domain Knowledge Fit in?Video + Language: Where Does Domain Knowledge Fit in?
Video + Language: Where Does Domain Knowledge Fit in?
 
Video + Language: Where Does Domain Knowledge Fit in?
Video + Language: Where Does Domain Knowledge Fit in?Video + Language: Where Does Domain Knowledge Fit in?
Video + Language: Where Does Domain Knowledge Fit in?
 
Csw2016 d antoine_automatic_exploitgeneration
Csw2016 d antoine_automatic_exploitgenerationCsw2016 d antoine_automatic_exploitgeneration
Csw2016 d antoine_automatic_exploitgeneration
 
An Introduction to Face Detection
An Introduction to Face DetectionAn Introduction to Face Detection
An Introduction to Face Detection
 
Detecting Food and Activities in Lifelogging Images
Detecting Food and Activities in Lifelogging ImagesDetecting Food and Activities in Lifelogging Images
Detecting Food and Activities in Lifelogging Images
 
Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3
Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3
Information to Wisdom: Commonsense Knowledge Extraction and Compilation - Part 3
 
Monitoring at Cloud Scale
Monitoring at Cloud ScaleMonitoring at Cloud Scale
Monitoring at Cloud Scale
 
A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015
A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015
A Standardized Encoding to Exchange Citizen Science Data - ESCA 2015
 
EDUPUB Implementation Demo Showcase - Reference SW using Readium JS
EDUPUB Implementation Demo Showcase - Reference SW using Readium JSEDUPUB Implementation Demo Showcase - Reference SW using Readium JS
EDUPUB Implementation Demo Showcase - Reference SW using Readium JS
 

Dernier

Davis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technologyDavis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technologycaarthichand2003
 
Pests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdfPests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdfPirithiRaju
 
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)Columbia Weather Systems
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensorsonawaneprad
 
Speech, hearing, noise, intelligibility.pptx
Speech, hearing, noise, intelligibility.pptxSpeech, hearing, noise, intelligibility.pptx
Speech, hearing, noise, intelligibility.pptxpriyankatabhane
 
RESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptx
RESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptxRESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptx
RESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptxFarihaAbdulRasheed
 
Pests of Bengal gram_Identification_Dr.UPR.pdf
Pests of Bengal gram_Identification_Dr.UPR.pdfPests of Bengal gram_Identification_Dr.UPR.pdf
Pests of Bengal gram_Identification_Dr.UPR.pdfPirithiRaju
 
Topic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxTopic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxJorenAcuavera1
 
THE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptx
THE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptxTHE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptx
THE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptxNandakishor Bhaurao Deshmukh
 
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptxSTOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptxMurugaveni B
 
Good agricultural practices 3rd year bpharm. herbal drug technology .pptx
Good agricultural practices 3rd year bpharm. herbal drug technology .pptxGood agricultural practices 3rd year bpharm. herbal drug technology .pptx
Good agricultural practices 3rd year bpharm. herbal drug technology .pptxSimeonChristian
 
Functional group interconversions(oxidation reduction)
Functional group interconversions(oxidation reduction)Functional group interconversions(oxidation reduction)
Functional group interconversions(oxidation reduction)itwameryclare
 
Pests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPirithiRaju
 
Transposable elements in prokaryotes.ppt
Transposable elements in prokaryotes.pptTransposable elements in prokaryotes.ppt
Transposable elements in prokaryotes.pptArshadWarsi13
 
GenBio2 - Lesson 1 - Introduction to Genetics.pptx
GenBio2 - Lesson 1 - Introduction to Genetics.pptxGenBio2 - Lesson 1 - Introduction to Genetics.pptx
GenBio2 - Lesson 1 - Introduction to Genetics.pptxBerniceCayabyab1
 
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...Universidade Federal de Sergipe - UFS
 
Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...
Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...
Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...lizamodels9
 
Base editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editingBase editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editingNetHelix
 

Dernier (20)

Davis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technologyDavis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technology
 
Pests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdfPests of safflower_Binomics_Identification_Dr.UPR.pdf
Pests of safflower_Binomics_Identification_Dr.UPR.pdf
 
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensor
 
Speech, hearing, noise, intelligibility.pptx
Speech, hearing, noise, intelligibility.pptxSpeech, hearing, noise, intelligibility.pptx
Speech, hearing, noise, intelligibility.pptx
 
RESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptx
RESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptxRESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptx
RESPIRATORY ADAPTATIONS TO HYPOXIA IN HUMNAS.pptx
 
Pests of Bengal gram_Identification_Dr.UPR.pdf
Pests of Bengal gram_Identification_Dr.UPR.pdfPests of Bengal gram_Identification_Dr.UPR.pdf
Pests of Bengal gram_Identification_Dr.UPR.pdf
 
Topic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxTopic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptx
 
THE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptx
THE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptxTHE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptx
THE ROLE OF PHARMACOGNOSY IN TRADITIONAL AND MODERN SYSTEM OF MEDICINE.pptx
 
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptxSTOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
STOPPED FLOW METHOD & APPLICATION MURUGAVENI B.pptx
 
Hot Sexy call girls in Moti Nagar,🔝 9953056974 🔝 escort Service
Hot Sexy call girls in  Moti Nagar,🔝 9953056974 🔝 escort ServiceHot Sexy call girls in  Moti Nagar,🔝 9953056974 🔝 escort Service
Hot Sexy call girls in Moti Nagar,🔝 9953056974 🔝 escort Service
 
Good agricultural practices 3rd year bpharm. herbal drug technology .pptx
Good agricultural practices 3rd year bpharm. herbal drug technology .pptxGood agricultural practices 3rd year bpharm. herbal drug technology .pptx
Good agricultural practices 3rd year bpharm. herbal drug technology .pptx
 
Functional group interconversions(oxidation reduction)
Functional group interconversions(oxidation reduction)Functional group interconversions(oxidation reduction)
Functional group interconversions(oxidation reduction)
 
Pests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdf
 
Transposable elements in prokaryotes.ppt
Transposable elements in prokaryotes.pptTransposable elements in prokaryotes.ppt
Transposable elements in prokaryotes.ppt
 
GenBio2 - Lesson 1 - Introduction to Genetics.pptx
GenBio2 - Lesson 1 - Introduction to Genetics.pptxGenBio2 - Lesson 1 - Introduction to Genetics.pptx
GenBio2 - Lesson 1 - Introduction to Genetics.pptx
 
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
 
Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...
Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...
Best Call Girls In Sector 29 Gurgaon❤️8860477959 EscorTs Service In 24/7 Delh...
 
Volatile Oils Pharmacognosy And Phytochemistry -I
Volatile Oils Pharmacognosy And Phytochemistry -IVolatile Oils Pharmacognosy And Phytochemistry -I
Volatile Oils Pharmacognosy And Phytochemistry -I
 
Base editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editingBase editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editing
 

Voginip lezing 2015: Classificeren zonder voorbeelden

  • 1. Visual Classification without examples Classificeren van beeld zonder voorbeelden Thomas Mensink VOGIN-IP-LEZING 2015
  • 2. • What is an axolotl? • Some examples Preview VOGIN-IP 20152
  • 3. VOGIN-IP 20153 We can classify based on labeled examples (supervised learning)
  • 4. Preview • What is an aye-aye? • Textual description: – Is nocturnal – Lives in trees – Has large eyes – Has long middle fingers VOGIN-IP 20154
  • 5. VOGIN-IP 20155 We can classify based on a textual description (and some prior knowledge)
  • 6. VOGIN-IP 20156 Can a computer do the same? (yes, that is what this talk is about)
  • 7. Agenda • Supervised Visual Classification • Attribute-Based Classification • Co-occurrence Based Classification VOGIN-IP 20157
  • 8. Computer Vision – in the news VOGIN-IP 20158
  • 10. Supervised Classification • Obtain annotated examples • Find a representation • Train a generic classifier VOGIN-IP 201510 Remarks: - New class: retrain on new examples - How to obtain training examples? - How to represent images?
  • 11. Visual Classification: Two trends Datasets • 2005: motorbikes, bicycles, people, cars • Since 2010: ImageNet 15K Representations • 2005: Manual derived features/encodings • 2012: Trained end-to- end, Deep Neural Nets VOGIN-IP 201511
  • 12. VOGIN-IP 201512 1000 classes, 5 guesses per image Current state-of-the-art: 6.7% error
  • 13. Estimating human performance VOGIN-IP 201513 Andrej Karpathy: “I realized that I needed to go through the painfully long training process myself”Test set 1500 images GoogleNet: 6.8% error Karpathy: 5.1% error
  • 14. VOGIN-IP 201514 Visual Classification: near-human performance when ample train data is available
  • 15. Attribute-based classification VOGIN-IP 201515 1. Define vocabulary 2. Train visual classifiers 3. Class to attribute mapping 4. Infer class
  • 16. What are good attributes? • Good attributes – are task and category dependent; – class discriminative, but not class specific; – interpretable by humans; and – detectable by computers VOGIN-IP 201516
  • 17. Quiz: What are good attributes? • is grey? • is made of atoms? • lives in Amsterdam? • is sunny? • eat fish? • has a SIFT descriptor with empty bin 3? • has 4 wheels? VOGIN-IP 201517
  • 18. How many attributes? • In theory k binary attributes can represent – 2k classes • In practice for c classes we need – Many attributes VOGIN-IP 201518
  • 20. Animals with Attributes - Vocabular VOGIN-IP 201520
  • 21. Class to attribute mapping VOGIN-IP 201521
  • 22. Attribute Based Prediction 1. Learn attribute classifiers from related classes 2. Train and Test set are disjoint 3. Infer attributes from new test image 4. Use attribute-to-class mapping to predict class VOGIN-IP 201522
  • 23. Animals with Attributes (results) VOGIN-IP 201523
  • 24. Disadvantages • Unnatural distinction between – Attributes to be detected – Classes of interest • Inherently multi-class zero-shot prediction VOGIN-IP 201524
  • 25. Classification based on co-occurrences I’m looking for a label, which I have not seen before. However, this picture contains also: 1. Indoor 2. Living room 3. Table 4. Chair VOGIN-IP 201525
  • 26. VOGIN-IP 201526 We can classify based on context
  • 28. COSTA: Design VOGIN-IP 201528 • Many visual concepts can be described as an open set of concept-to-concept relations • Describe image semantics with co-occurrences • Exploit natural bias in natural images
  • 29. Exploit natural bias in natural images VOGIN-IP 201529 Sink"is"u visual"spa a"stove," (2) onal to the ments weestimate lab truthlabellingof our i occurrencestatisticsc pora,e.g.,Wordnet or suchasYahoo, Google 3 Lreg = i kwi − k akw = X i X d wid − a> whereindex i and k both run over th sii = 0. The vector vid contains weightedweight vectorsvidk = sik w Notethat thelossisformulatedove over train images. Moreover, Eq. (9) = i d wid − whereindex i and k both run ov sii = 0. The vector vid conta weightedweight vectorsvidk = s Notethat thelossisformulated over train images. Moreover, Eq obtainedinclosed-formusingrid weobservethat regularization is formance, thedimensionality of a (1) etween the s paper, we atistics be- erent simi- two labels. informativeclueabout thev isalsoshowninanimagere In addition to the positi c++ ij , wealsousetheother p thepresenceof label i with senceof label i withthepres of bothlabels, denotedby c+ i eachof thesedefinitionsof larity measuresdefinedabov Using the positive and weight vector w of anunkn
  • 30. COSTA: Classifier • Goal: Estimate classifier for unseen label • Knowledge base: – k trained classifiers – Co-occurrences • Zero-shot classifier: VOGIN-IP 201530
  • 31. Co-Occurrence Statistics • Ground-truth data (proof-of-concept) • Web search engines • Flickr Tags • Language resources • Visual annotated data (eg Microsoft COCO) VOGIN-IP 201531
  • 32. Example: Beach Holiday VOGIN-IP 201532 Concept Normalized Co-Oc Weight Sea 0.1810 Water 0.0992 Summer 0.0548 LandscapeNature 0.0435 SunsetSunrise 0.0383 Sports 0.0367 Travel 0.0347 Ship 0.0346 Sunny 0.0319 Big Group 0.0282
  • 35. Co-occurrences from the Web VOGIN-IP 201535
  • 36. Conclusions • Supervised visual classification performs well when ample train data is available • Classification without examples: – Define some set of base classifiers – Transfer new class to space of these classifiers – Two examples: attributes and co-occurrences VOGIN-IP 201536
  • 37. Thanks to: • The organizers • Christoph Lampert for slides and inspiration • Authors of the cited papers • Colleagues and supervisors (UvA: Amir, Cees, Jan, Spencer & Stratis, PhD: Cordelia, Florent, Gabriela, Jakob) VOGIN-IP 201537
  • 38. Literature • Frome, Corrado, Shlens, Bengio, Dean, Ranzato,and Mikolov, “DeViSE: A Deep Visual-Semantic Embedding Model”, NIPS 2013 • Habibian, Mensink, and Snoek, “VideoStory: A New Multimedia Embedding for Few-Example Recognition and Translation of Events”, ACM MM 2014 • Lampert, Nickish, and Harmeling, “Attribute-Based Classification for Zero-Shot Learning of Object Categories”, TPAMI 2013 • Li, Gavves, Mensink, and Snoek, “Attributes Make Sense on Segmented Objects”, ECCV 2014 • Mensink, Gavves, and Snoek, “COSTA: Co-Occurrence Statistics for Zero-Shot Classification”, CVPR 2014 • Norouzi, Mikolov, Bengio, Singer, Shlens, Frome, Corrado, and Dean, “Zero-Shot Learning by Convex Combination of Semantic Embeddings”, ICLR 2014 VOGIN-IP 201538

Notes de l'éditeur

  1. 9