Image Tag Clarity: In Search of Visual- Representative Tags for Social Images

Similar documents
PRISM: Concept-preserving Social Image Search Results Summarization

TagProp: Discriminative Metric Learning in Nearest Neighbor Models for Image Auto-Annotation

Setting up Flickr. Open a web browser such as Internet Explorer and type this url in the address bar.

The little mermaid lives in a beautiful castle in the deep blue sea. She lives with her five sisters and her father, the Merking.

Social Media Modeling and Computing

Tree structured CRF models for interactive image labeling

TagProp: Discriminative Metric Learning in Nearest Neighbor Models for Image Annotation

Welcome to the class of Web Information Retrieval. Min ZHANG

Lab for Media Search, National University of Singapore 1

AFS-World & City package

Object Recognition I. Linda Shapiro EE/CSE 576

Universities Access IBM/Google Cloud Compute Cluster for NSF-Funded Research

Small particles scatter light

Contents. Contents. Perfecting people shots Making your camera a friend.5. Beyond point and shoot Snapping to the next level...

A Text-Based Approach to the ImageCLEF 2010 Photo Annotation Task

Using Queried Keywords or Full-text Extracted Keywords in Blog Mining?

BUAA AUDR at ImageCLEF 2012 Photo Annotation Task

Fab Investment Outlook Foundry, Memory and LED. Clark Tseng, Sr. Research Manager, SEMI Vietnam Semiconductor Strategy Summit September 16-17, 20214

Chapter 5- Setting Up a World

The Fraunhofer IDMT at ImageCLEF 2011 Photo Annotation Task

SkyFinder: Attribute-based Sky Image Search

Available Light Photographic Cheat Sheet v1.90 main reference tables

A Consumer Photo Tagging Ontology

Image Annotation in Presence of Noisy Labels

Using Tags for Instant Access to Photos with Snapper.Photo s PhotoManager Welcome to the World of S napper.photo

LSP 121. LSP 121 Math and Tech Literacy II. Topics. Quartiles. Intro to Statistics. More Descriptive Statistics

From Pixels to Semantics Mining Digital Imagery Data for Automatic Linguistic Indexing of Pictures

Using Coherence-based Measures to Predict Query Difficulty

Promoting Disaster Reduction through Multi- National Cooperation in Asia region. ADRC Activities. Jun 2009 Asian Disaster Reduction Center (ADRC)

Multimedia Information Retrieval The case of video

IEEE 1857 Standard Empowering Smart Video Surveillance Systems

Setting Up Your Environment - Consider the width of your camera view. Large waves may look good from afar but when zoomed in often look too large.

Course Outline. Microsoft SharePoint Server 2013 for the Site Owner/Power User Course 55035: 2 days Instructor-Led

MIchael Hudson. Photography Over 100 New Images

Using Temporal Profiles of Queries for Precision Prediction

WallArt_Inner_V1.pdf 1 6/29/ :56:13 PM

ACM MM Dong Liu, Shuicheng Yan, Yong Rui and Hong-Jiang Zhang

FLAMINGO CHEAT SHEET FOR ES 305 CLASS 5 Pages of Fresh Flamingo Goodness

Object Class Recognition using Images of Abstract Regions

About LIDAR Data. What Are LIDAR Data? How LIDAR Data Are Collected

Associating video frames with text

The Growing Problem of Mobile Adware

Science & Technology Group

Characterization and Modeling of Deleted Questions on Stack Overflow

Retrieval and Feedback Models for Blog Distillation

State Water Survey Division

You ll also be able to add markers like Placemarks to keep track of some special locations.

MULTIMODAL SEMI-SUPERVISED IMAGE CLASSIFICATION BY COMBINING TAG REFINEMENT, GRAPH-BASED LEARNING AND SUPPORT VECTOR REGRESSION

Computer with Microsoft PowerPoint and Internet access STUDENT WORKSHEET: Making a Presentation with Microsoft PowerPoint

HYPERVARIATE DATA VISUALIZATION

Natural Language Processing

Joint Inference in Image Databases via Dense Correspondence. Michael Rubinstein MIT CSAIL (while interning at Microsoft Research)

EUROPEAN KANGOUROU LINGUISTICS ENGLISH-LEVELS 3-4. Linguistic ENGLISH. LEVEL: 3 4 (Γ - Δ Δημοτικού)

Time-aware Approaches to Information Retrieval

Auto Flash Off Portrait Landscape Action

IATF Stakeholder Conference

ICT Year 5. Unit 5A: Writing for different audiences

Introduction to SharePoint 2013 for Collaboration and Document Management

(and what the numbers mean)

Class 5: Attributes and Semantic Features

CSC 110 Lab 12 Graphics and Objects. Names:

QLogic/Lenovo 16Gb Gen 5 Fibre Channel for Database and Business Analytics

NOWPAP. Northwest Pacific Action Plan. United Nations Environment Programme

CYBERTECH MIDWEST Indianapolis, Indiana

over Multi Label Images

Visual Dictionary: Towards a Higher-level Visual Representation for Object Categorization. CHUA, Tat-Seng School of Computing

Document and Query Expansion Models for Blog Distillation

AVANTUS TRAINING PTE PTE LTD LTD

Comparison of Feature Sets using Multimedia Translation

Advanced Topics in Information Retrieval. Learning to Rank. ATIR July 14, 2016

Retrieval and Feedback Models for Blog Distillation

Guernsey Post 2013/14. Quality of Service Report

Daytime Long Exposures

Chapter 9. Attribute joins

QLogic 16Gb Gen 5 Fibre Channel for Database and Business Analytics

Lenses & Exposure. Lenses. Exposure. Lens Options Depth of Field Lens Speed Telephotos Wide Angles. Light Control Aperture Shutter ISO Reciprocity

Visuals Distributed Specializing in Management & Delivery of Hospitality Visuals

Fish species recognition from video using SVM classifier

NUS-WIDE: A Real-World Web Image Database from National University of Singapore

Course Outline. Module 1: SharePoint Overview

Tag Based Image Search by Social Re-ranking

Exploiting noisy web data for largescale visual recognition

Understanding the use of Temporal Expressions on Persian Web Search

Writing Reports with Report Designer and SSRS 2014 Level 1

easy english readers The Little Mermaid By Hans Christian Andersen

Automated Online News Classification with Personalization

Blog Site Search Using Resource Selection

MIT805 BIG DATA MAPREDUCE

Annotation and Evaluation

Photo Tourism: Exploring Photo Collections in 3D

DUTH at ImageCLEF 2011 Wikipedia Retrieval

24 hours of Photo Sharing. installation by Erik Kessels

2014 Google Earth Engine Research Award Report

Developing Focused Crawlers for Genre Specific Search Engines

Visual Query Suggestion

INTERNET MARKETING May 2009

Unit 5.A Properties of Light Essential Fundamentals of Light 1. Electromagnetic radiation has oscillating magnetic and electric components.

PMA Goes Global. The MARPA Program for Promoting PMA Around the World

Section 9: One Variable Statistics

Procedural Generation of Videos to Train Deep Action Recognition Networks (Supplementary Material)

Transcription:

Image Tag Clarity: In Search of Visual- Representative Tags for Social Images Aixin Sun, Sourav S. Bhowmick Nanyang Technological University Singapore 1

Outline Web search & query clarity Image tag search and clarity Experiments and evaluation Conclusion and discussion 2

Web search examples: bank vs. 2008 3

Summarized by Wordle.net: bank vs. 2008 4

Query performance prediction in web search Query is effective: the retrieved documents contain unusually large probabilities of words specific to the topic Query is not effective: the retrieved documents is similar to a set of randomly sampled documents the word probability distribution is similar to that of the collection Query clarity score [Cronen-Townsend 02]: 5

Tag is visually representative? The images associated with the tag are visually similar to each other It is relatively easy to find a small set of images representing the tagged images (or the tag). Tag: Sunset Tag: Zebra Tag: Asia? Tag: 2008? 6

Sunset, Zebra, Asia, 2008 from Flickr.com Sunset Asia Zebra 2008 7

Finding visual-representative tag Query: a tag t Retrieved documents: all images annotated with the tag T Image representation: bag of visual-words (as documents) Image tag clarity score: 8

Tag language model All images (documents) are equally important Images closer to the centroid are more important [Elsas 08] 9

Tag language model vs. collection language model 10

And we have a little problem here 11

Expected tag clarity score: derived from randomly assigned dummy tags 12

Normalized image tag clarity score Given a tag t with frequency freq(t) Compute the expected tag clarity score and standard derivation with multiple dummy tags of the same frequency Normalized tag clarity through zero-mean normalization Approximation: bin frequency 13

Experiments on NUS-WIDE Dataset NUS-WIDE http://lms.comp.nus.edu.sg/research/nus-wide.htm Features 500D bag of visual-words Images 269,648 Tags 5981 (with frequency >=100) Categories (or concepts) All 81 category labels appear as tags in the dataset 14

Nitc distribution 15

Top-50 Most Visual-representative Tags Tag Nitc Pfreq Tag Nitc Pfreq Tag Nitc Pfreq sunset 319.6 99.67 minimal 104.6 77.08 airplane 78.7 96.72 silhouette 211.2 97.64 beach 104.3 99.41 sand 78.4 98.08 fog 207.5 96.71 dunes 100.9 82.59 cloud 77.5 97.61 sky 197.6 99.97 dawn 100.5 91.09 foggy 77.1 68.18 sunrise 179.2 97.76 ocean 100.2 99.03 weather 76.5 95.47 charts 158.1 78.36 moon 100 94.87 morning 75.7 96.64 sun 151.9 99.1 lake 98.9 98.5 pattern 74.2 92.63 mist 138.6 94.75 night 94.1 99.5 atardecer 74.1 71.96 clouds 133.9 99.85 graphs 94 6.89 jet 74.1 93.56 lightning 129.5 73.72 graph 91.3 1.97 lines 73.7 94.9 blue 118.1 99.95 longexposure 91 97.71 dusk 73.4 95.13 sea 116.3 99.52 zebra 89.8 82.46 moleskine 72.8 70.76 minimalism 114.9 77.66 chart 89.6 20.7 southcascades 71.5 6.02 landscape 110.2 99.77 sketches 87.9 81.52 water 70.4 99.93 windmills 106.7 75.76 plane 83.8 95.79 unbuilding 70 67.31 storm 106 96.22 aircraft 82.4 96.2 craterlakenationalpark 69.4 10.58 horizon 105.5 92.44 seascape 80.6 91.72 16

Top-50 Least Visual-representative Tags Tag Nitc Pfreq Tag Nitc Pfreq Tag Nitc Pfreq people -2.9 99.55 august -1.3 80.12 june -1 86.52 asia -2.5 98.26 photographers -1.3 85.74 pics -1 58.08 brown -2.4 96.87 finepix -1.3 65.36 bottle -1 55.63 japan -2.3 98.11 religion -1.2 94.67 april -1 84.53 washington -2.2 97.06 photos -1.2 94.22 september -1 75.66 2008-2.1 98.51 smorgasbord -1.2 61.33 hungary -1 62.82 france -2 98.39 panasonic -1.2 85.32 caribou -1 80.77 picture -1.7 92.49 global -1.2 65.74 cannon -1 58.23 photograph -1.6 88.86 may -1.1 83.51 or -1 24.08 july -1.6 86.42 israel -1.1 86.94 exotic -1 62.18 china -1.6 96.99 outside -1.1 92.81 lumix -1 86.57 virginia -1.5 86.99 cool -1.1 95.7 republic -1 37.07 india -1.5 97.44 culture -1.1 93.13 canadian -0.9 64.62 ohio -1.3 87.33 royal -1.1 72.71 this -0.9 41.87 maryland -1.3 84.17 world -1.1 95.34 prayer -0.9 85.72 colorful -1.3 97.53 2005-1.1 96.05 persian -0.9 64.04 pic -1.3 58.7 iranian -1 57.33 17

Sunset vs People 18

Distribution of the 81 category labels 46 are highly representative with nitc >=10 26 are representative with 2<=nitc<10; 9 (or 11%) are non-representative with nitc<2. Events and activities: dancing, running, soccer, sports, earthquake, Scene and location: castle, town, house, temple. 19

Tag Frequency vs Tag Clarity Tag frequency percentile (bin) 20

Conclusion and discussion The concept of clarity Web search and query clarity Tag search and tag clarity The computation of tag visual-representativeness Expected tag clarity through dummy tags Normalized tag clarity score Evaluation of on NUS-Wide dataset Future work Applications: tag recommendation? image classification? Computation: sampling of tagged images? Global features? Evaluation: is water more visually representative than river? 21

Acknowledgement Lab for Media Search, for sharing the NUS-WIDE dataset MSRA grant for partially supporting this trip References: S. Cronen-Townsend, Y. Zhou, W. B. Croft: Predicting query performance. SIGIR 2002: 299-306 J. L. Elsas, J. Arguello, J. Callan, J. G. Carbonell: Retrieval and feedback models for blog feed search. SIGIR 2008: 347-354 22