What Helps Where – And Why? Semantic Relatedness for Knowledge Transfer Marcus Rohrbach 1,2 Michael Stark 1,2 György Szarvas 1 Iryna Gurevych 1 Bernt Schiele.

Slides:

Advertisements

Similar presentations

Recognizing Human Actions by Attributes CVPR2011 Jingen Liu, Benjamin Kuipers, Silvio Savarese Dept. of Electrical Engineering and Computer Science University.

Advertisements

Google Similarity Distance Presented by: Akshay Kumar Pankaj Prateek.

Large-Scale Object Recognition using Label Relation Graphs Jia Deng 1,2, Nan Ding 2, Yangqing Jia 2, Andrea Frome 2, Kevin Murphy 2, Samy Bengio 2, Yuan.

Human Action Recognition by Learning Bases of Action Attributes and Parts Bangpeng Yao, Xiaoye Jiang, Aditya Khosla, Andy Lai Lin, Leonidas Guibas, and.

UCB Computer Vision Animals on the Web Tamara L. Berg CSE 595 Words & Pictures.

Problem Semi supervised sarcasm identification using SASI

Performance Evaluation Measures for Face Detection Algorithms Prag Sharma, Richard B. Reilly DSP Research Group, Department of Electronic and Electrical.

Query Dependent Pseudo-Relevance Feedback based on Wikipedia SIGIR ‘09 Advisor: Dr. Koh Jia-Ling Speaker: Lin, Yi-Jhen Date: 2010/01/24 1.

Discriminative Segment Annotation in Weakly Labeled Video Kevin Tang, Rahul Sukthankar Appeared in CVPR 2013 (Oral)

GENERATING AUTOMATIC SEMANTIC ANNOTATIONS FOR RESEARCH DATASETS AYUSH SINGHAL AND JAIDEEP SRIVASTAVA CS DEPT., UNIVERSITY OF MINNESOTA, MN, USA.

Enhancing Exemplar SVMs using Part Level Transfer Regularization 1.

Self Taught Learning : Transfer learning from unlabeled data Presented by: Shankar B S DMML Lab Rajat Raina et al, CS, Stanford ICML 2007.

Relative Attributes Presenter: Shuai Zheng (Kyle) Supervised by Philip H.S. Torr Author: Devi Parikh (TTI-Chicago) and Kristen Grauman (UT-Austin)

Hierarchical Subquery Evaluation for Active Learning on a Graph Oisin Mac Aodha, Neill Campbell, Jan Kautz, Gabriel Brostow CVPR 2014 University College.

Ordinal Decision Trees Qinghua Hu Harbin Institute of Technology

Improving web image search results using query-relative classifiers Josip Krapacy Moray Allanyy Jakob Verbeeky Fr´ed´eric Jurieyy.

Jinhui Tang †, Shuicheng Yan †, Richang Hong †, Guo-Jun Qi ‡, Tat-Seng Chua † † National University of Singapore ‡ University of Illinois at Urbana-Champaign.

School of Electronic Information Engineering, Tianjin University Human Action Recognition by Learning Bases of Action Attributes and Parts Jia pingping.

Thumbs Up or Thumbs Down? Semantic Orientation Applied to Unsupervised Classification on Reviews Peter D. Turney Institute for Information Technology National.

Query Relevance Feedback and Ontologies How to Make Queries Better.

Query Expansion.

Human abilities Presented By Mahmoud Awadallah 1.

Machine Learning Approach for Ontology Mapping using Multiple Concept Similarity Measures IEEE/ACIS International Conference on Computer and Information.

Studying Relationships Between Human Gaze, Description, and Computer Vision Kiwon Yun 1, Yifan Peng 1 Dimitris Samaras 1, Gregory J. Zelinsky 1,2, Tamara.

Combining Lexical Semantic Resources with Question & Answer Archives for Translation-Based Answer Finding Delphine Bernhard and Iryna Gurevvch Ubiquitous.

Marcin Marszałek, Ivan Laptev, Cordelia Schmid Computer Vision and Pattern Recognition, CVPR Actions in Context.

PAUL ALEXANDRU CHIRITA STEFANIA COSTACHE SIEGFRIED HANDSCHUH WOLFGANG NEJDL 1* L3S RESEARCH CENTER 2* NATIONAL UNIVERSITY OF IRELAND PROCEEDINGS OF THE.

GA-Based Feature Selection and Parameter Optimization for Support Vector Machine Cheng-Lung Huang, Chieh-Jen Wang Expert Systems with Applications, Volume.

A Word at a Time: Computing Word Relatedness using Temporal Semantic Analysis Kira Radinsky (Technion) Eugene Agichtein (Emory) Evgeniy Gabrilovich (Yahoo!

Presenter: Lung-Hao Lee ( 李龍豪 ) January 7, 309.

The More the Better? Assessing the Influence of Wikipedia’s Growth on Semantic Relatedness Measures Torsten Zesch and Iryna Gurevych Ubiquitous Knowledge.

80 million tiny images: a large dataset for non-parametric object and scene recognition CS 4763 Multimedia Systems Spring 2008.

Enhancing Human-Machine Communication via Visual Attributes Devi Parikh Virginia Tech.

Reduction of Training Noises for Text Classifiers Rey-Long Liu Dept. of Medical Informatics Tzu Chi University Taiwan.

Semantic Embedding Space for Zero Shot Action Recognition Xun XuTimothy HospedalesShaogang GongAuthors: Computer Vision Group Queen Mary University of.

IEEE Int'l Symposium on Signal Processing and its Applications 1 An Unsupervised Learning Approach to Content-Based Image Retrieval Yixin Chen & James.

Transfer Learning for Image Classification Group No.: 15 Group member : Feng Cai Sauptik Dhar Sauptik.

Binxing Jiao et. al (SIGIR ’10) Presenter : Lin, Yi-Jhen Advisor: Dr. Koh. Jia-ling Date: 2011/4/25 VISUAL SUMMARIZATION OF WEB PAGES.

Comparing and Ranking Documents Once our search engine has retrieved a set of documents, we may want to Rank them by relevance –Which are the best fit.

Sharing Features Between Objects and Their Attributes Sung Ju Hwang 1, Fei Sha 2 and Kristen Grauman 1 1 University of Texas at Austin, 2 University of.

A Scalable Machine Learning Approach for Semi-Structured Named Entity Recognition Utku Irmak(Yahoo! Labs) Reiner Kraft(Yahoo! Inc.) WWW 2010(Information.

CIKM Opinion Retrieval from Blogs Wei Zhang 1 Clement Yu 1 Weiyi Meng 2 1 Department of.

Visual Categorization With Bags of Keypoints Original Authors: G. Csurka, C.R. Dance, L. Fan, J. Willamowski, C. Bray ECCV Workshop on Statistical Learning.

1/21 Automatic Discovery of Intentions in Text and its Application to Question Answering (ACL 2005 Student Research Workshop )

Prepared by: Mahmoud Rafeek Al-Farra College of Science & Technology Dep. Of Computer Science & IT BCs of Information Technology Data Mining

Improving Search Results Quality by Customizing Summary Lengths Michael Kaisser ★, Marti Hearst  and John B. Lowe ★ University of Edinburgh,  UC Berkeley,

Gang WangDerek HoiemDavid Forsyth. INTRODUCTION APROACH (implement detail) EXPERIMENTS CONCLUSION.

Flickr Tag Recommendation based on Collective Knowledge BÖrkur SigurbjÖnsson, Roelof van Zwol Yahoo! Research WWW Summarized and presented.

1 A Web Search Engine-Based Approach to Measure Semantic Similarity between Words Presenter: Guan-Yu Chen IEEE Trans. on Knowledge & Data Engineering,

2015/12/121 Extracting Key Terms From Noisy and Multi-theme Documents Maria Grineva, Maxim Grinev and Dmitry Lizorkin Proceeding of the 18th International.

Data Mining, ICDM '08. Eighth IEEE International Conference on Duy-Dinh Le National Institute of Informatics Hitotsubashi, Chiyoda-ku Tokyo,

Image Classification over Visual Tree Jianping Fan Dept of Computer Science UNC-Charlotte, NC

ACADS-SVMConclusions Introduction CMU-MMAC Unsupervised and weakly-supervised discovery of events in video (and audio) Fernando De la Torre.

Comparing Word Relatedness Measures Based on Google n-grams Aminul ISLAM, Evangelos MILIOS, Vlado KEŠELJ Faculty of Computer Science Dalhousie University,

Towards Total Scene Understanding: Classiﬁcation, Annotation and Segmentation in an Automatic Framework N 工科所錢雅馨 2011/01/16 Li-Jia Li, Richard.

11 A Classification-based Approach to Question Routing in Community Question Answering Tom Chao Zhou 1, Michael R. Lyu 1, Irwin King 1,2 1 The Chinese.

Describing People: A Poselet-Based Approach to Attribute Classification.

Identifying “Best Bet” Web Search Results by Mining Past User Behavior Author: Eugene Agichtein, Zijian Zheng (Microsoft Research) Source: KDD2006 Reporter:

Semantic search-based image annotation Petra Budíková, FI MU CEMI meeting, Plzeň,

NEIL: Extracting Visual Knowledge from Web Data Xinlei Chen, Abhinav Shrivastava, Abhinav Gupta Carnegie Mellon University CS381V Visual Recognition -

Bangpeng Yao1, Xiaoye Jiang2, Aditya Khosla1,

Machine Learning overview Chapter 18, 21

Mining the Data Charu C. Aggarwal, ChengXiang Zhai

Basic Intro Tutorial on Machine Learning and Data Mining

Quanzeng You, Jiebo Luo, Hailin Jin and Jianchao Yang

Prepared by: Mahmoud Rafeek Al-Farra

Web Page Cleaning for Web Mining

Using Uneven Margins SVM and Perceptron for IE

Zeroshot Learning Mun Jonghwan.

Using Link Information to Enhance Web Page Classification

Presentation transcript:

What Helps Where – And Why? Semantic Relatedness for Knowledge Transfer Marcus Rohrbach 1,2 Michael Stark 1,2 György Szarvas 1 Iryna Gurevych 1 Bernt Schiele 1,2 1 Department of Computer Science, TU Darmstadt 2 MPI Informatics, Saarbrücken

2 Knowledge transfer for zero-shot object class recognition CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Group classes by attributes [Lampert et al., CVPR `09] Manual supervision: Object class-attribute associations Group classes by attributes [Lampert et al., CVPR `09] Manual supervision: Object class-attribute associations Describing using attributes [Farhadi et al., CVPR `09 & `10] Manual supervision: Attribute labels Describing using attributes [Farhadi et al., CVPR `09 & `10] Manual supervision: Attribute labels animal four legged mammal white paw Unseen class (no training images) Giant panda ? Attributes for knowledge transfer

3 Knowledge transfer for zero-shot object class recognition CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Describing using attributes [Farhadi et al., CVPR `09 & `10] Manual supervision: Attribute labels Describing using attributes [Farhadi et al., CVPR `09 & `10] Manual supervision: Attribute labels animal four legged mammal Group classes by attributes [Lampert et al., CVPR `09] Manual supervision: Object class-attribute associations Group classes by attributes [Lampert et al., CVPR `09] Manual supervision: Object class-attribute associations white paw Unseen class (no training images) Giant panda ? Attributes for knowledge transfer  Replace manual supervision b y semantic relatedness mined from language resources  Unsupervised Transfer WordNet Attributes for knowledge transfer

4 Attribute-based model [Lampert et al., CVPR `09] CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | ocean spots … … Known training classes Attribute classifiers Unseen test classes Class-attribute associations [Lampert et al., CVPR `09] Supervised: manual (human judges) Attributes white

5 Attribute-based model [Lampert et al., CVPR `09] CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language [Lampert et al., CVPR `09] Supervised: manual (human judges) ocean spots … … Known training classes Attribute classifiers Unseen test classes Class-attribute associations white

6 Direct similarity-based model CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language ocean spots … … Known training classes Attribute classifiers Unseen test classes Class-attribute associations white

7 Direct similarity-based model CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language Known training classes Unseen test classes Class-attribute associations Classifier per class killer whale Dalmatian polar bear

8 Direct similarity-based model CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language Unseen test classes most similar classes Known training classes Classifier per class polar bear killer whale Dalmatian … …

9  Models for visual knowledge transfer  Semantic relatedness measures  Language resources  WordNet  Wikipedia  WWW  Image search  Respective state-of-the-art measures  Evaluation  Conclusion Outline CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

10  WordNet  Lin measure [Budanitsky & Hirst, CL `06]  Wikipedia  Explicit Semantic Analysis [Gabrilovich & MarkovitchI, IJCAI `07]  Word Wide Web  Hitcount (Dice coeffient) [Kilgarriff & Grefenstette, CL `03]  Image search  Visually more relevant  Hitcount (Dice coeffient) Semantic Relatedness Measures CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | WordNet [Fellbaum, MIT press `98] WordNet [Fellbaum, MIT press `98] entityanimalmammalhorseelephantvehiclecarbike

11  WordNet  Lin measure [Budanitsky & Hirst, CL `06]  Wikipedia  Explicit Semantic Analysis [Gabrilovich & MarkovitchI, IJCAI `07]  Word Wide Web  Hitcount (Dice coeffient) [Kilgarriff & Grefenstette, CL `03]  Image search  Visually more relevant  Hitcount (Dice coeffient) Semantic Relatedness Measures CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | WordNet [Fellbaum, MIT press `98] WordNet [Fellbaum, MIT press `98] entityanimalmammalhorseelephantvehiclecarbike

12  WordNet  Lin measure [Budanitsky & Hirst, CL `06]  Wikipedia  Explicit Semantic Analysis [Gabrilovich & MarkovitchI, IJCAI `07]  Word Wide Web  Hitcount (Dice coeffient) [Kilgarriff & Grefenstette, CL `03]  Image search  Visually more relevant  Hitcount (Dice coeffient) Semantic Relatedness Measures CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | A farm is an area of lan the training of horses. A hoof is the tip of a toe Rear hooves of a horse Hoof Farm Tusks are long teeth, u Elephants and narwhals Tusk Articlehorseelephant Farm30 Hoof21 Tusk04 ……… A farm is an area of lan the training of horses. A hoof is the tip of a toe Rear hooves of a horse Most evem tped ungulat Hoof Farm Tusks are long teeth, u Elephants and narwhals Tusk

13  WordNet  Lin measure [Budanitsky & Hirst, CL `06]  Wikipedia  Explicit Semantic Analysis [Gabrilovich & MarkovitchI, IJCAI `07]  Word Wide Web  Hitcount (Dice coeffient) [Kilgarriff & Grefenstette, CL `03]  Image search  Visually more relevant  Hitcount (Dice coeffient) Semantic Relatedness Measures CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | A farm is an area of lan the training of horses. A hoof is the tip of a toe Rear hooves of a horse Hoof Farm Tusks are long teeth, u Elephants and narwhals Tusk Articlehorseelephant Farm30 Hoof21 Tusk04 ……… A farm is an area of lan the training of horses. A hoof is the tip of a toe Rear hooves of a horse Most evem tped ungulat Hoof Farm Tusks are long teeth, u Elephants and narwhals Tusk cosine

14  WordNet  Lin measure [Budanitsky & Hirst, CL `06]  Wikipedia  Explicit Semantic Analysis [Gabrilovich & MarkovitchI, IJCAI `07]  Word Wide Web  Hitcount (Dice coeffient) [Kilgarriff & Grefenstette, CL `03]  Image search  Visually more relevant  Hitcount (Dice coeffient) Semantic Relatedness Measures CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

15  WordNet  Lin measure [Budanitsky & Hirst, CL `06]  Wikipedia  Explicit Semantic Analysis [Gabrilovich & MarkovitchI, IJCAI `07]  Word Wide Web  Hitcount (Dice coeffient) [Kilgarriff & Grefenstette, CL `03]  Image search  Visually more relevant  Hitcount (Dice coeffient) Semantic Relatedness Measures CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | We watched a horse race yesterday. [..] Tomorrow we go in the zoo to look at the baby elephant. „the dance of the horse and elephant“ web searchimage search [ lahierophant/ /] Incidental co-occurence Terms refer to same entity (the image)

16  Models for visual knowledge transfer  Semantic relatedness measures  Evaluation  Attributes  Querying class-attribute associations  Mining attributes  Direct similarity  Attribute-based vs. direct similarity  Conclusion Outline CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

17  Animals with attributes dataset [Lampert et al., CVPR `09]  40 training, 10 test classes (disjoint)  ≈ images total  Downsampled to 92 training images per class  Manual associations to 85 attributes  Image classification  SVM: Histogram intersection kernel  Area under ROC curve (AUC) - chance level: 50%  Mean over all 10 test classes Experimental Setup CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

18 Performance of supervised approach CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

19 Querying: abbreviation agile Manual supervision: detailed description “having a high degree of physical coordination” Querying: abbreviation agile Manual supervision: detailed description “having a high degree of physical coordination”  Performance of queried association  Encouraging  Below manual supervision  Image search  Based on image related text  Wikipedia  Robust resource  Yahoo Web  Very noisy resource  WordNet  Path length poor indicator of class-attribute associations Querying class-attribute association CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

20  Performance of queried association  Encouraging  Below manual supervision  Image search (Yahoo Img, Flickr)  Based on image related text  Wikipedia  Robust resource (definition texts)  Yahoo Web  Very noisy resource  WordNet  Path length poor indicator of class-attribute associations Querying class-attribute association CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | the dance of the horse and elephant image search

21  Performance of queried association  Encouraging  Below manual supervision  Image search (Yahoo Img, Flickr)  Based on image related text  Wikipedia  Robust resource (definition text)  Yahoo Web  Very noisy resource  WordNet  Path length poor indicator of class-attribute associations Querying class-attribute association CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Noise: While he watched a horse race the leg of his chair broke. Noise: While he watched a horse race the leg of his chair broke.

22  Performance of queried association  Encouraging  Below manual supervision  Image search (Yahoo Img, Flickr)  Based on image related text  Wikipedia  Robust resource (definition text)  Yahoo Web  Very noisy resource  WordNet  Path length poor indicator of class-attribute associations Querying class-attribute association CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

23 Mining attributes CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language Attribute terms ocean spots … … Known training classes Unseen test classes Class-attribute associations white

24 Mining attributes CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language Attribute terms ? ? ? ? ? ? Known training classes Unseen test classes Class-attribute associations ? ?

25 Part attributes Leg of a dog Attribute classifiers Mining attributes CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language Known training classes Unseen test classes Class-attribute associations flipperlegpaw Word Net

26  Additional measure: Holonym patterns  Only part attributes  Hit Counts of Patterns [Berland & Charniak, ACL 1999]  “cow’s leg”  “leg of a cow”  Dice coefficient Mining attributes CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | While he watched a horse race the leg of his chair broke. Leg of the horse web searchholonym patterns Incidental co-occurence One term likely part of other term

27  Best: Yahoo Holonyms  Close to manual attributes  Tailored towards part attributes Mining attributes CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

28  Best: Yahoo Holonyms  Close to manual attributes  Tailored towards part attributes  Performance drop  Reduced diversity  Only part attributes  Specialized terms  E.g. pilus (=hair)  Coverage problem: Image search, Wikipedia Mining attributes CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

29  Models for visual knowledge transfer  Semantic relatedness measures  Evaluation  Attributes  Querying class-attribute associations  Mining attributes  Direct similarity  Attribute-based vs. direct similarity  Conclusion Outline CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

30 Direct similarity-based model CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Word Net semantic relatedness from language Unseen test classes most similar classes Known training classes Classifier per class polar bear killer whale Dalmatian

31  Nearly all very good  On par with manual supervision attribute model (black)  Clearly better than any mined attribute-associations result  Why?  Five most related classes  Ranking of semantic relatedness reliable  Similar between methods Direct Similarity CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

32 Attributes vs. direct similarity CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |  Extending the test set  Add images  From known classes  As negatives  More realistic setting  Results  Direct similarity drop in performance (orange curve)  Attribute models generalize well

33  Models for visual knowledge transfer  Semantic relatedness measures  Evaluation  Conclusion Outline CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

34  Supervision replaced with semantic relatedness  Direct similarity  better than attributes  on par with supervised approach  Attributes: generalizes better  Semantic relatedness measures  Overall best  Yahoo image with hit count  Holonym patterns for web search  Improvement  Limited to part attributes Conclusion CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

35  Supervision replaced with semantic relatedness  Direct similarity  better than attributes  on par with supervised approach  Attributes: generalize better  Semantic relatedness measures  Overall best  Yahoo image with hit count  Holonym patterns for web search  Improvement  Limited to part attributes Conclusion CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach |

36  Supervision replaced with semantic relatedness  Direct similarity  better than attributes  on par with supervised approach  Attributes: generalize better  Semantic relatedness measures  Overall best  Yahoo image with hit count  Holonym patterns for web search  Improvement  Limited to part attributes  WordNet poor for object-attributes associations Conclusion CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | patterns: dog’s leg leg of the dogs patterns: dog’s leg leg of the dogs

37 Further supervision for closing the semantic gap? See us at our poster (A2, Atrium)! CVPR 2010 | What Helps Where - And Why? Semantic Relatedness for Knowledge Transfer | Marcus Rohrbach | Knowledge Transfer Thank you! Software?