SlideShare a Scribd company logo
1 of 55
SherlockNet:
Tagging and Captioning the British Library
1 Million Image Dataset
Brian Do, Karen Wang, Luda Zhao
The SherlockNet Team
Luda
BS in CS,
MS in CS @
Stanford
Brian
BS in CS @ Stanford,
MD-PhD student @
Harvard/MIT
Karen
BS in CS,
MS in CS @
Stanford
One Million Images
Problem: Current Tags Not Very Useful
Solution: Use Neural Networks to Generate
Descriptive Tags For Every Image
building
architecture
temple
pyramid
stone
etc...
Roadmap
- What are Neural Networks?
- The First Pass: Category Classification
- Lots of Tags: OCR Tagging Using Related Images
- Image Captioning
- Presenting: SherlockNet Interface
What Are Neural Networks?
Problem: Classifying Inputs Into Categories
Photo Credits: Andrej Karpathy, CS 231N, Stanford University
The Input is Passed Through Multiple Layers
Photo Credits: Andrej Karpathy, CS 231N, Stanford University
Convolutional Neural Networks (CNNs) are
Optimized for Images as Input
Photo Credits: Andrej Karpathy, CS 231N, Stanford University
Tagging
We trained a CNN to classify all 1M images into
one of 12 categories
people: 0.80
architecture: 0.12
diagrams: 0.05
object: 0.02
decoration: 0.01
Confidence
percentage
We trained a CNN to classify all 1M images into
one of 12 categories
81% top-1 accuracy
97% top-3 accuracy
We trained a CNN to classify all 1M images into
one of 12 categories
We trained a CNN to classify all 1M images into
one of 12 categories
We trained a CNN to classify all 1M images into
one of 12 categories
Advanced Tagging
Can we use surrounding text to
get better tags (Google Images approach)?
number
writers
broods
season
sparrow
hedge
bird
species
colouring
breast
Optical
Character
Recognition
Can we use surrounding text to
get better tags (Google Images approach)?
person
top
head
cheeks
spot
ring
part
back
rump
throat
We “vectorize” images and minimize Euclidean
distance to obtain related images
CNN
<3,4,-1,-3,4>
We “vectorize” images and minimize Euclidean
distance to obtain related images
CNN
<3,4,-1,-3,4>
<2,4,-1,-5,3>
<-3,2,5,3,-3>
<-1,0,0,5,1>
<3,3,0,-3,5>
D = 6
D = 161
D = 106
D = 3
We “vectorize” images and minimize Euclidean
distance to obtain related images
CNN
<3,4,-1,-3,4>
<2,4,-1,-5,3>
<-3,2,5,3,-3>
<-1,0,0,5,1>
<3,3,0,-3,5>
D = 6
D = 161
D = 106
D = 3
birds
trees
london
park
stick
planting
wing
claws
beaks
nuts
wings
pacific
species
people
rainbow
pairing
birder
park
wing
descriptions
species
trees
beak
london
perching
eyes
bird
benches
We pooled surrounding text from similar images
bird
tree
london
park
stick
plant
wing
claws
beak
nuts
wing
pacific
species
people
rainbow
pair
bird
park
wing
description
species
tree
beak
london
perch
eye
bird
bench
We then “stemmed” words + did spell check
We then had similar images “vote” on tags
bird
tree
london
park
stick
plant
wing
claws
beak
nuts
wing
pacific
species
people
rainbow
pair
bird
park
wing
description
species
tree
beak
london
perch
eye
bird
bench
We pooled surrounding text from similar images
bird
park
wing
species
beak
+ =
This makes the tags for each image much cleaner
and more refined
We pooled surrounding text from similar images
We pooled surrounding text from similar images
We pooled surrounding text from similar images
We pooled surrounding text from similar images
Captioning
Motivation
- Most natural way of showcasing
images
- Opportunities to provide
contextual information, “the man
next to the woman”
- From AI research standpoint:
interesting theoretical challenges
Background
- Combining two distinct
neural networks (CNNs and
RNNs) to do end-to-end
processing
- Very active area of research
Challenges
- High quality photographs vs.
low-res, black & white
illustrations
- Ambiguity in detail levels
- Difficulty in obtaining ground
truth data(“machines can’t
learn without prior knowledge”)
vs
.
???
Worked well...
Almost there...
Not quite sure what’s happening
A New Dataset
- British Museum Prints and Drawings Collection
- Over ~200,000 images through public interface
- Many have good, human-annotated captions
- Potential for machine learning research?
(From www.britishmuseum.org Online catalogue)
Preliminary Results
Training tags from captions:
.
.
.
Bridges
Ships
We can extract informative tags from our captions
SherlockNet Web Interface
bit.ly/sherlocknet
SherlockNet will one day provide multiple levels of
high-quality text annotation for every image
Tags: Architecture, landscape, river, trees, boat
Caption: A boat on a tree-lined river in front of a building
Acknowledgements
The British Library
Mahendra Mahey
Adam Farquhar
Hana Lewis
Adrian Edwards
Elliot Crowley
Mario Klingemann
Ben O'Steen
Stanford University
Andrej Karpathy
Justin Johnson
Stefano Ermon
The British Museum
SherlockNet will one day provide multiple levels of
high-quality text annotation for every image
Tags: Architecture, landscape, river, trees, boat
Caption: A boat on a tree-lined river in front of a building
Stylistic Trends Over Time
Machine Detection of Style
Neural networks reveal image features that become
more or less frequent over time
Feature #541 is highly activated in modern decorations compared to antique decorations
Neural networks reveal image features that become
more or less frequent over time
Images with high
score for Feature #541
Images with low
score for Feature #541
Feature #541 is highly activated in modern decorations compared to antique decorations
Neural networks reveal image features that become
more or less frequent over time
Images with high
score for Feature #541
Images with low
score for Feature #541
Feature #541 is highly activated in modern decorations compared to antique decorations
Feature #541 probably indicates the presence of lines
delineating the top and bottom of the decoration.
Process + Results
- Decorations: 64% accuracy, Maps: 52% accuracy (compared to 16%/
20% accuracy random chance, respectively)
- Pretty good results given inherent limitations!
SherlockNet’s contact details
Karen Wang - kwang37@stanford.edu
Luda Zhao – ludazhao@stanford.edu
Brian Do - brian_do@hms.harvard.edu

More Related Content

Similar to SherlockNet

Interpretation, Context, and Metadata: Examples from Open Context
Interpretation, Context, and Metadata: Examples from Open ContextInterpretation, Context, and Metadata: Examples from Open Context
Interpretation, Context, and Metadata: Examples from Open ContextEric Kansa
 
Neo4j - graph database for recommendations
Neo4j - graph database for recommendationsNeo4j - graph database for recommendations
Neo4j - graph database for recommendationsproksik
 
Multimedia Data Navigation and the Semantic Web (SemTech 2006)
Multimedia Data Navigation and the Semantic Web (SemTech 2006)Multimedia Data Navigation and the Semantic Web (SemTech 2006)
Multimedia Data Navigation and the Semantic Web (SemTech 2006)Bradley Allen
 
Neo4j graphdatabaseforrecommendations-130531021030-phpapp02-converted
Neo4j graphdatabaseforrecommendations-130531021030-phpapp02-convertedNeo4j graphdatabaseforrecommendations-130531021030-phpapp02-converted
Neo4j graphdatabaseforrecommendations-130531021030-phpapp02-convertedsnehapandey01
 
Natural Language Processing in R (rNLP)
Natural Language Processing in R (rNLP)Natural Language Processing in R (rNLP)
Natural Language Processing in R (rNLP)fridolin.wild
 
MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...
MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...
MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...Rob Lancefield
 
Digital Medieval Data Curation
Digital Medieval Data CurationDigital Medieval Data Curation
Digital Medieval Data Curationblalbritton
 
Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...
Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...
Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...Facultad de Informática UCM
 
Civil War Data 150 at DLF Fall Forum 2011
Civil War Data 150 at DLF Fall Forum 2011Civil War Data 150 at DLF Fall Forum 2011
Civil War Data 150 at DLF Fall Forum 2011Jon Voss
 
Hub and Spokes Development June07
Hub and Spokes Development June07Hub and Spokes Development June07
Hub and Spokes Development June07Jane Stevenson
 
Ch03 Mining Massive Data Sets stanford
Ch03 Mining Massive Data Sets  stanfordCh03 Mining Massive Data Sets  stanford
Ch03 Mining Massive Data Sets stanfordSakthivel C R
 
'E-Science and Archaeology'
'E-Science and Archaeology''E-Science and Archaeology'
'E-Science and Archaeology'Stuart Dunn
 
Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)
Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)
Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)Bradley Allen
 
Alberto Massidda - Images and words: mechanics of automated captioning with n...
Alberto Massidda - Images and words: mechanics of automated captioning with n...Alberto Massidda - Images and words: mechanics of automated captioning with n...
Alberto Massidda - Images and words: mechanics of automated captioning with n...Codemotion
 
Query Translation for Ontology-extended Data Sources
Query Translation for Ontology-extended Data SourcesQuery Translation for Ontology-extended Data Sources
Query Translation for Ontology-extended Data SourcesJie Bao
 
Anchor Modeling Introduction Med Res
Anchor Modeling Introduction Med ResAnchor Modeling Introduction Med Res
Anchor Modeling Introduction Med ResRoenbaeck
 
ALIAOnline Practical Linked (Open) Data for Libraries, Archives & Museums
ALIAOnline Practical Linked (Open) Data for Libraries, Archives & MuseumsALIAOnline Practical Linked (Open) Data for Libraries, Archives & Museums
ALIAOnline Practical Linked (Open) Data for Libraries, Archives & MuseumsJon Voss
 
An Open Context for Archaeology
An Open Context for ArchaeologyAn Open Context for Archaeology
An Open Context for Archaeologyguest756e05
 

Similar to SherlockNet (20)

Interpretation, Context, and Metadata: Examples from Open Context
Interpretation, Context, and Metadata: Examples from Open ContextInterpretation, Context, and Metadata: Examples from Open Context
Interpretation, Context, and Metadata: Examples from Open Context
 
Neo4j - graph database for recommendations
Neo4j - graph database for recommendationsNeo4j - graph database for recommendations
Neo4j - graph database for recommendations
 
Multimedia Data Navigation and the Semantic Web (SemTech 2006)
Multimedia Data Navigation and the Semantic Web (SemTech 2006)Multimedia Data Navigation and the Semantic Web (SemTech 2006)
Multimedia Data Navigation and the Semantic Web (SemTech 2006)
 
Neo4j graphdatabaseforrecommendations-130531021030-phpapp02-converted
Neo4j graphdatabaseforrecommendations-130531021030-phpapp02-convertedNeo4j graphdatabaseforrecommendations-130531021030-phpapp02-converted
Neo4j graphdatabaseforrecommendations-130531021030-phpapp02-converted
 
Natural Language Processing in R (rNLP)
Natural Language Processing in R (rNLP)Natural Language Processing in R (rNLP)
Natural Language Processing in R (rNLP)
 
MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...
MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...
MCN 2013: Open Access Image Delivery at the Davison Art Center, Wesleyan Univ...
 
Digital Medieval Data Curation
Digital Medieval Data CurationDigital Medieval Data Curation
Digital Medieval Data Curation
 
Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...
Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...
Analyzing large multimedia collections in an urban context - Prof. Marcel Wor...
 
Civil War Data 150 at DLF Fall Forum 2011
Civil War Data 150 at DLF Fall Forum 2011Civil War Data 150 at DLF Fall Forum 2011
Civil War Data 150 at DLF Fall Forum 2011
 
Hub and Spokes Development June07
Hub and Spokes Development June07Hub and Spokes Development June07
Hub and Spokes Development June07
 
Ch03 Mining Massive Data Sets stanford
Ch03 Mining Massive Data Sets  stanfordCh03 Mining Massive Data Sets  stanford
Ch03 Mining Massive Data Sets stanford
 
'E-Science and Archaeology'
'E-Science and Archaeology''E-Science and Archaeology'
'E-Science and Archaeology'
 
Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)
Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)
Searching BBC Rushes Using Semantic Web Techniques (TRECVID 2005)
 
Alberto Massidda - Images and words: mechanics of automated captioning with n...
Alberto Massidda - Images and words: mechanics of automated captioning with n...Alberto Massidda - Images and words: mechanics of automated captioning with n...
Alberto Massidda - Images and words: mechanics of automated captioning with n...
 
Query Translation for Ontology-extended Data Sources
Query Translation for Ontology-extended Data SourcesQuery Translation for Ontology-extended Data Sources
Query Translation for Ontology-extended Data Sources
 
Matching Dirty Data
Matching Dirty DataMatching Dirty Data
Matching Dirty Data
 
Anchor Modeling Introduction Med Res
Anchor Modeling Introduction Med ResAnchor Modeling Introduction Med Res
Anchor Modeling Introduction Med Res
 
ALIAOnline Practical Linked (Open) Data for Libraries, Archives & Museums
ALIAOnline Practical Linked (Open) Data for Libraries, Archives & MuseumsALIAOnline Practical Linked (Open) Data for Libraries, Archives & Museums
ALIAOnline Practical Linked (Open) Data for Libraries, Archives & Museums
 
Esad 12may2010
Esad 12may2010Esad 12may2010
Esad 12may2010
 
An Open Context for Archaeology
An Open Context for ArchaeologyAn Open Context for Archaeology
An Open Context for Archaeology
 

More from labsbl

7th BL Labs Symposium (2019): 13_Closing comments
7th BL Labs Symposium (2019): 13_Closing comments7th BL Labs Symposium (2019): 13_Closing comments
7th BL Labs Symposium (2019): 13_Closing commentslabsbl
 
7th BL Labs Symposium (2019): 12_Digital Research team projects update
7th BL Labs Symposium (2019): 12_Digital Research team projects update7th BL Labs Symposium (2019): 12_Digital Research team projects update
7th BL Labs Symposium (2019): 12_Digital Research team projects updatelabsbl
 
7th BL Labs Symposium (2019): 11_The Artistic Award
7th BL Labs Symposium (2019): 11_The Artistic Award7th BL Labs Symposium (2019): 11_The Artistic Award
7th BL Labs Symposium (2019): 11_The Artistic Awardlabsbl
 
7th BL Labs Symposium (2019): 10_British Library Staff Award
7th BL Labs Symposium (2019): 10_British Library Staff Award7th BL Labs Symposium (2019): 10_British Library Staff Award
7th BL Labs Symposium (2019): 10_British Library Staff Awardlabsbl
 
7th BL Labs Symposium (2019): 09_Community commendation
7th BL Labs Symposium (2019): 09_Community commendation7th BL Labs Symposium (2019): 09_Community commendation
7th BL Labs Symposium (2019): 09_Community commendationlabsbl
 
7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project
7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project
7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ projectlabsbl
 
7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...
7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...
7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...labsbl
 
7th BL Labs Symposium (2019): 05_The Research Award
7th BL Labs Symposium (2019): 05_The Research Award7th BL Labs Symposium (2019): 05_The Research Award
7th BL Labs Symposium (2019): 05_The Research Awardlabsbl
 
7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...
7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...
7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...labsbl
 
7th BL Labs Symposium (2019): 03_BL Labs update
7th BL Labs Symposium (2019): 03_BL Labs update7th BL Labs Symposium (2019): 03_BL Labs update
7th BL Labs Symposium (2019): 03_BL Labs updatelabsbl
 
7th BL Labs Symposium (2019): 01_Welcome and Introduction
7th BL Labs Symposium (2019): 01_Welcome and Introduction7th BL Labs Symposium (2019): 01_Welcome and Introduction
7th BL Labs Symposium (2019): 01_Welcome and Introductionlabsbl
 
7th BL Labs Symposium (2019): 07_The Teaching & Learning Award
7th BL Labs Symposium (2019): 07_The Teaching & Learning Award7th BL Labs Symposium (2019): 07_The Teaching & Learning Award
7th BL Labs Symposium (2019): 07_The Teaching & Learning Awardlabsbl
 
Digital Magical Mystery Tour - British Library
Digital Magical Mystery Tour - British LibraryDigital Magical Mystery Tour - British Library
Digital Magical Mystery Tour - British Librarylabsbl
 
Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...
Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...
Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...labsbl
 
Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...
Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...
Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...labsbl
 
Bl labs sfu-dhi_lab-dhilab-2019-workshop
Bl labs sfu-dhi_lab-dhilab-2019-workshopBl labs sfu-dhi_lab-dhilab-2019-workshop
Bl labs sfu-dhi_lab-dhilab-2019-workshoplabsbl
 
Building Better GLAM Labs - Keynote Presentation at Simon Fraser University
Building Better GLAM Labs - Keynote Presentation at Simon Fraser UniversityBuilding Better GLAM Labs - Keynote Presentation at Simon Fraser University
Building Better GLAM Labs - Keynote Presentation at Simon Fraser Universitylabsbl
 
Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion Project ...
Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion  Project ...Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion  Project ...
Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion Project ...labsbl
 
BL Labs Presentation at Open Science Infrastructures for Big Cultural Data
BL Labs Presentation at Open Science Infrastructures for Big Cultural DataBL Labs Presentation at Open Science Infrastructures for Big Cultural Data
BL Labs Presentation at Open Science Infrastructures for Big Cultural Datalabsbl
 
A hands-on data exploration & challenge to become a derived data-set author o...
A hands-on data exploration & challenge to become a derived data-set author o...A hands-on data exploration & challenge to become a derived data-set author o...
A hands-on data exploration & challenge to become a derived data-set author o...labsbl
 

More from labsbl (20)

7th BL Labs Symposium (2019): 13_Closing comments
7th BL Labs Symposium (2019): 13_Closing comments7th BL Labs Symposium (2019): 13_Closing comments
7th BL Labs Symposium (2019): 13_Closing comments
 
7th BL Labs Symposium (2019): 12_Digital Research team projects update
7th BL Labs Symposium (2019): 12_Digital Research team projects update7th BL Labs Symposium (2019): 12_Digital Research team projects update
7th BL Labs Symposium (2019): 12_Digital Research team projects update
 
7th BL Labs Symposium (2019): 11_The Artistic Award
7th BL Labs Symposium (2019): 11_The Artistic Award7th BL Labs Symposium (2019): 11_The Artistic Award
7th BL Labs Symposium (2019): 11_The Artistic Award
 
7th BL Labs Symposium (2019): 10_British Library Staff Award
7th BL Labs Symposium (2019): 10_British Library Staff Award7th BL Labs Symposium (2019): 10_British Library Staff Award
7th BL Labs Symposium (2019): 10_British Library Staff Award
 
7th BL Labs Symposium (2019): 09_Community commendation
7th BL Labs Symposium (2019): 09_Community commendation7th BL Labs Symposium (2019): 09_Community commendation
7th BL Labs Symposium (2019): 09_Community commendation
 
7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project
7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project
7th BL Labs Symposium (2019): 08_An update on the ‘Living with machines’ project
 
7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...
7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...
7th BL Labs Symposium (2019): 06_An overview of digital preservation at the B...
 
7th BL Labs Symposium (2019): 05_The Research Award
7th BL Labs Symposium (2019): 05_The Research Award7th BL Labs Symposium (2019): 05_The Research Award
7th BL Labs Symposium (2019): 05_The Research Award
 
7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...
7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...
7th BL Labs Symposium (2019): 04_The story of the GLAM Labs community and how...
 
7th BL Labs Symposium (2019): 03_BL Labs update
7th BL Labs Symposium (2019): 03_BL Labs update7th BL Labs Symposium (2019): 03_BL Labs update
7th BL Labs Symposium (2019): 03_BL Labs update
 
7th BL Labs Symposium (2019): 01_Welcome and Introduction
7th BL Labs Symposium (2019): 01_Welcome and Introduction7th BL Labs Symposium (2019): 01_Welcome and Introduction
7th BL Labs Symposium (2019): 01_Welcome and Introduction
 
7th BL Labs Symposium (2019): 07_The Teaching & Learning Award
7th BL Labs Symposium (2019): 07_The Teaching & Learning Award7th BL Labs Symposium (2019): 07_The Teaching & Learning Award
7th BL Labs Symposium (2019): 07_The Teaching & Learning Award
 
Digital Magical Mystery Tour - British Library
Digital Magical Mystery Tour - British LibraryDigital Magical Mystery Tour - British Library
Digital Magical Mystery Tour - British Library
 
Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...
Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...
Building Better GLAM Labs - Opening talk at Museum Big Data Conference - UCL ...
 
Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...
Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...
Building Better GLAM Labs - Keynote at University of Victoria, Victoria, BC, ...
 
Bl labs sfu-dhi_lab-dhilab-2019-workshop
Bl labs sfu-dhi_lab-dhilab-2019-workshopBl labs sfu-dhi_lab-dhilab-2019-workshop
Bl labs sfu-dhi_lab-dhilab-2019-workshop
 
Building Better GLAM Labs - Keynote Presentation at Simon Fraser University
Building Better GLAM Labs - Keynote Presentation at Simon Fraser UniversityBuilding Better GLAM Labs - Keynote Presentation at Simon Fraser University
Building Better GLAM Labs - Keynote Presentation at Simon Fraser University
 
Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion Project ...
Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion  Project ...Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion  Project ...
Introduction to BL Labs and Reading 35,000 Books: The UCD Contagion Project ...
 
BL Labs Presentation at Open Science Infrastructures for Big Cultural Data
BL Labs Presentation at Open Science Infrastructures for Big Cultural DataBL Labs Presentation at Open Science Infrastructures for Big Cultural Data
BL Labs Presentation at Open Science Infrastructures for Big Cultural Data
 
A hands-on data exploration & challenge to become a derived data-set author o...
A hands-on data exploration & challenge to become a derived data-set author o...A hands-on data exploration & challenge to become a derived data-set author o...
A hands-on data exploration & challenge to become a derived data-set author o...
 

Recently uploaded

Solving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptxSolving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptxOH TEIK BIN
 
KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...
KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...
KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...M56BOOKSTORE PRODUCT/SERVICE
 
Accessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactAccessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactdawncurless
 
Introduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxIntroduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxpboyjonauth
 
Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Educationpboyjonauth
 
How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17Celine George
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)eniolaolutunde
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfsanyamsingh5019
 
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxiammrhaywood
 
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Krashi Coaching
 
MENTAL STATUS EXAMINATION format.docx
MENTAL     STATUS EXAMINATION format.docxMENTAL     STATUS EXAMINATION format.docx
MENTAL STATUS EXAMINATION format.docxPoojaSen20
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Crayon Activity Handout For the Crayon A
Crayon Activity Handout For the Crayon ACrayon Activity Handout For the Crayon A
Crayon Activity Handout For the Crayon AUnboundStockton
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptxVS Mahajan Coaching Centre
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionSafetyChain Software
 
The basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxThe basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxheathfieldcps1
 
Paris 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityParis 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityGeoBlogs
 

Recently uploaded (20)

Solving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptxSolving Puzzles Benefits Everyone (English).pptx
Solving Puzzles Benefits Everyone (English).pptx
 
Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝
Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝
Model Call Girl in Tilak Nagar Delhi reach out to us at 🔝9953056974🔝
 
KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...
KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...
KSHARA STURA .pptx---KSHARA KARMA THERAPY (CAUSTIC THERAPY)————IMP.OF KSHARA ...
 
Accessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactAccessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impact
 
Staff of Color (SOC) Retention Efforts DDSD
Staff of Color (SOC) Retention Efforts DDSDStaff of Color (SOC) Retention Efforts DDSD
Staff of Color (SOC) Retention Efforts DDSD
 
Introduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxIntroduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptx
 
Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Education
 
How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdf
 
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
 
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
 
MENTAL STATUS EXAMINATION format.docx
MENTAL     STATUS EXAMINATION format.docxMENTAL     STATUS EXAMINATION format.docx
MENTAL STATUS EXAMINATION format.docx
 
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdfTataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Crayon Activity Handout For the Crayon A
Crayon Activity Handout For the Crayon ACrayon Activity Handout For the Crayon A
Crayon Activity Handout For the Crayon A
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory Inspection
 
The basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxThe basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptx
 
Paris 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityParis 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activity
 

SherlockNet

Editor's Notes

  1. We each introduce ourselves.
  2. Talk about the British Library’s Flickr Commons collections. Contains more than a million images from the British Library’s digitized collection of over 65,000 books. 15th - 19th centuries Subjects: Literature, Science, Anthropology, many more Put online by the British Library for researchers and the public to use them in novel and interesting ways.
  3. Date Volume Page Current tags not very useful
  4. Talk about neural networks class. Convolutional neural networks, or CNNs for short. This bleeding-edge computer vision technology has been used in the past couple of years to perform image recognition with extremely successful results.. ...even outperforming humans! Final project goal: use neural networks to generate descriptive tags for every image in the British Library Flickr collection.
  5. In our project we used CNNs to classify each image into a category find related images and generate captions. CNNs are very, very good at the above tasks. We’ll talk briefly about why this is, and how they wor.
  6. At a high level, a neural network takes an input and, for each possible category, it computes a score. A higher score means the input is more likely to be in that category.
  7. In the process of computing the scores, the input is passed through multiple layers. At each layer, the neural network is “activated” by features of increasing levels of complexity. These activations are determined by parameters that the neural network learns over time. The concept is analogous to the activation of biological neurons that form the communication network of our brain and spinal cord. The multiple layers of a neural network allow it to recognize complex models - the more neurons (computational power), the more complex the models it can recognize, and thus better classification results.
  8. Convolutional neural networks are specialized for images as input. Because images follow a images have width, height, and depth, CNNs are able to optimize for this architecture by only computing activations for a small region of the input. For example, it can look for localized features like an edge or a blotch of color. I.e. can recognize a visual feature that appears in multiple places, facing different ways, at different angles and sizes, etc. The name, convolution, comes from the mathematical operation that is performed between the input and the neural network’s parameters at each layer. Tie this into Brian’s 12 categories.
  9. Explain SherlockNet Labs Signpost what we’re going to do: Next two sections we are calling them “SherlockNet Labs”, for tasks that we think are a little bit beyond what’s currently feasible with neural nets, but this projects allowed us the opportunity to explore them a bit, and our hope is that we can inspire further research in these topics
  10. Explain SherlockNet Labs Signpost what we’re going to do: Next two sections we are calling them “SherlockNet Labs”, for tasks that we think are a little bit beyond what’s currently feasible with neural nets, but this projects allowed us the opportunity to explore them a bit, and our hope is that we can inspire further research in these topics
  11. Why do we need/want captions? most natural ways of showcasing
  12. Around 5% to 10% of the dataset
  13. Two years ago I studied abroad at Oxford. There I did a tutorial on the history of British architecture. One thing I found really challenging was trying to find records of architecture in books. I spent hours sitting in the Bodleian hunting down books. Oxford, architecture
  14. To recap, we have tagged over 1 Million images with convolutional neural networks and generated hundred of thousands of human-readable captions for them, while providing an interface for people to explore them easily. Our ultimate hope for this project is to have it serve as a prototype for any digital collection to make their content more discoverable, dramatically cut down on years of manual labelling, and provide tools to discover deeper insights into the rich, rich materials that history has left us.
  15. We want to thank all of our mentor and collaborators, especially the British Labs and Mahendra Mahey for working with us through multiple time zones to make this project happen. It’s been great fun. Thank you!