DOCSLIB.ORG

Samy Bengio Real Life Setting

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Samy Bengio Real Life Setting Sharing is Caring in the Land of The Long Tail Samy Bengio Real life setting “Real problems rarely come packaged as 1M images uniformly belonging to a set of 1000 classes…” 2 The long tail • Well known phenomena where a small number of generic objects/entities/words appear very often and most others appear more rarely. • Also knows as Zipf or Power law, or Pareto distribution. • The web is littered by this kind of distributions: • the frequency of each unique query on search engines, • the occurrences of each unique word in text documents, • etc. 3 Example of a long tail Frequency of words in Wikipedia 1e+09 1e+08 1e+07 1e+06 100000 Frequency 10000 1000 100 10 the anyways trickiest h-plane Words 4 Representation sharing • How do we design a classifier or a ranker when data follows a long tail distribution? • If we train one model per class, it is hard for poor classes to be well trained. • How come we humans are able to recognize objects we have seen only once or even never? • Most likely answer: representation sharing: all class models share/learn a joint representation. • Poor classes can then benefit from knowledge learned from semantically similar but richer classes. • Extreme case: zero-shot setting! 5 Outline In this talk, I will cover the following ideas: • Wsabie: a joint embedding space of images and labels • The many facets of text embeddings • Zero-shot setting through embeddings • Incorporate Knowledge Graph constraints • Use of a language model I will NOT cover the following important issues: • Prediction time issues for extreme classification • Memory issues 6 Wsabie Learn to embed images & labels to optimize top-ranked items. Labels Obama Eiffel Tower Shark Dolphin Lion ... 100-dimensional embedding space Wsabie: J. Weston et al, ECML 2010, IJCAI 2011 7 Wsabie: summary sim(i,x) = <Wi,Vx> W V 000100 real values Label i Image x Triplet Loss: sim( , dolphin) > sim( , obama) + 1 Trained by stochastic gradient descent and smart sampling of negative examples 8 Wsabie: experiments - results Method ImageNet 2010 Web prec@1 prec@10 prec@1 prec@10 approx 1.55% 0.41% 0.30% 0.34% kNN One-vs- 2.27% 1.02% 0.52% 0.29% Rest Wsabie 4.03% 1.48% 1.03% 0.44% Ensemble of 10 10.03% 3.02% Wsabies ImageNet 2010: 16000 labels and 4M images Web: 109000 labels and 16M images 9 Wsabie: embeddings Label Nearest Neighbors barack obama barak obama, obama, barack, barrack obama, bow wow david beckham beckham, david beckam, alessandro del piero, del piero santa santa claus, papa noel, pere noel, santa clause, joyeux noel dolphin delphin, dauphin, whale, delfin, delfini, baleine, blue whale cows cattle, shire, dairy cows, kuh, horse, cow, shire horse, kone rose rosen, hibiscus, rose flower, rosa, roze, pink rose, red rose eiffel tower eiffel, tour eiffel, la tour eiffel, big ben, paris, blue mosque ipod i pod, ipod nano, apple ipod, ipod apple, new ipod f18 f 18, eurofighter, f14, fighter jet, tomcat, mig 21, f 16 10 Wsabie: annotations delfini, orca, dolphin, mar, delfin, dauphin, whale, cancun, killer whale, sea world blue whale, whale shark, great white shark, underwater, white shark, shark, manta ray, dolphin, requin, blue shark, diving barrack obama, barak obama, barack hussein obama, barack obama, james marsden, jay z, obama, nelly, falco, barack eiffel, paris by night, la tour eiffel, tour eiffel, eiffel tower, las vegas strip, eifel, tokyo tower, eifel tower 11 “Why not an embedding of text only?” Skip-Gram (Word2Vec) Learn dense embedding vectors from an unannotated text corpus, e.g. Wikipedia wing chair Obama W tuna E bull shark “an exceptionally large male tiger shark can grow up to” tiger shark http://code.google.com/p/word2vec Tomas Mikolov, Kai Chen, Greg Corrado, Jeff Dean (ICLR 2013) 13 Skip-Gram Wikipedia t-SNE visualization of ImageNet Skip-gram trained on Wikipedia, labels 155K terms tiger shark car bull shark cars blacktip shark muscle car shark sports car oceanic whitetip shark compact car sandbar shark autocar dusky shark automobile blue shark pickup truck requiem shark racing car great white shark passenger car reptiles lemon shark dealership birds insects food musical instruments clothing dogs aquatic life animals transportation 14 Embeddings are powerful Berlin hot Germany big Rome hotter Italy bigger E(Rome) - E(Italy) + E(Germany) ≈ E(Berlin) E(hotter) - E(hot) + E(big) ≈ E(bigger) 15 Let’s go back to images! Deep convolutional models for images But what about the long tail of classes? Layer 7 ... What about using our semantic embeddings Layer 1 for that? Input 17 ConSE: Convex Combination of Semantic Embeddings [Norouzi et al, ICLR’2014] p(Lion x) | p(Apple x) | p(Orange x) | p(Tiger x) | p(Bear x) | 18 ConSE: Convex Combination of Semantic Embeddings from Skip-Gram for instance: s(y) = embedding position of y f(x)= p(y x)s(y ) i| i i f(x)=pX(Lion x)s(Lion)+ p(Apple| x)s(Apple)+ p(Orange| x)s(Orange)+ | p(Tiger x)s(Tiger)+ | p(Bear x)s(Bear) | Do a nearest neighbor search around f(x) to find the corresponding label 19 ConSE(T): Convex Combination of Semantic Embeddings In practice, consider the average of only a few labels: top(T )= i p(y x) is among top T probabilities { | i| } 1 f(x)= p(y x)s(y ) Z i| i i top(T ) 2X 20 ConSE(T): experiments on ImageNet • Model trained with 1.2M 3-hops ILSVRC 2012 images from 1,000 classes 2-hops • Evaluated on images from same classes. Training • Results are measured as hit@k. ConSe(T) experiments 22 Knowledge Graph 23 Multiclass Classifiers Softmax GoogleLeNet model Logistic 24 Object labels have rich relations Exclusion Hierarchical Dog Dog Cat Corgi Puppy Cat Corgi Puppy Overlap 25 Visual Model + Knowledge Graph Dog 0.9 Visual Knowledge Corgi 0.8 Model Joint Inference Graph Puppy 0.9 Cat 0.1 Hierarchy and Exclusion (HEX) Graph Exclusion Hierarchical Dog Cat [Deng et al, ECCV 2014] Corgi Puppy 26 HEX Classification Model x ∈ Rn y ∈ {0,1}n Input scores Binary Label vector 1 (y, y ) Pr(y| x) = ∏φi (xi, yi ) ∏ψi, j i j Z(x) i i, j sigmoid(x ) if y =1 0 If violates constraints φ (x , y ) = i i ψ (y, y ) = i i i 1 ( ) if y 0 i, j i j − sigmoid xi i = 1 Otherwise Unary: same as logistic regression Pairwise: set illegal configuration to zero ! All illegal configurations have probability zero. 27 Exp: Learning with weak labels • ILSVRC 2012: “relabel” or “weaken” a portion of fine-grained leaf labels to basic level labels. • Evaluate on fine-grained recognition Animal Animal Animal Relabel Dog … Dog Dog … … Corgi Husky Corgi Husky Corgi Husky Original ILSVRC 2012 Training Test (leaf labels) (“weakened” labels) 28 Exp: Learning with weak labels • ILSVRC 2012: “relabel” or “weaken” a portion of fine-grained leaf labels to basic level labels. • Evaluate on fine-grained recognition. • Consistently outperforms baselines. Top 1 accuracy (top 5 accuracy) 29 What about textual descriptions? • We have considered the long tail of objets. • What about more complex descriptions, involving multi-word descriptions, or captions? • We can use language models to help. 30 Neural Image Caption Generator [Vinyals et al, CVPR 2015] Vision Language 1. Two pizzas sitting on top of Deep CNN Generating a stove top RNN oven. 2. A pizza sitting on top of a pan on top of a stove. A group of people Vision! Language ! Deep CNN Generating! shopping at an RNN outdoor market. ! There are many vegetables at the fruit stand. 31 NIC: objective • Let I be an image (pixels). • Let S be the corresponding sentence (sequence of words). • Likelihood of producing the right sentence given the image: N log p(S I)= log p(St I,S0,...,St 1) | | − t=0 X • We maximize the likelihood of producing the right sentence given the image: ✓? = arg max log p(S I; ✓) ✓ | (XI,S) 32 NIC: model P(word 1) P(word 2) P(<end>) Embedding word 1 word N Recurrent Image Convolution Neural Net Neural Net 33 Examples A person riding a A skateboarder does a trick A dog is jumping to catch a motorcycle on a dirt road. Two dogs play in the grass. on a ramp. frisbee. A group of young people Two hockey players are A little girl in a pink hat is A refrigerator filled with lots of playing a game of frisbee. fighting over the puck. blowing bubbles. food and drinks. A herd of elephants walking A close up of a cat laying A red motorcycle parked on the A yellow school bus parked across a dry grass field. on a couch. side of the road. in a parking lot. Describes without errors Describes with minor errors Somewhat related to the image Unrelated to the image 34 It doesn’t always work… Human: A blue and black dress ... No! I see white and gold! Our model: A close up of a vase with flowers. Scheduled Sampling[NIPS 2015] 36 Scheduled Sampling 37 Scheduled Sampling 1 0.9 Exponential decay Inverse sigmoid decay 0.8 Linear decay 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 200 400 600 800 1000 38 Conclusions • The long tail problem happens in most of the interesting tasks. • Sharing approaches can help “poor” classes to generalize thanks to “rich” classes. • At the extreme, semantic embeddings can represent classes with zero training examples. • Sharing approaches are interesting not only for text and images, but also for complete sentences.
Load more

Recommended publications
  • Read Book Thierry Henry : Lonely at The
    THIERRY HENRY : LONELY AT THE TOP PDF, EPUB, EBOOK Philippe Auclair | 352 pages | 18 Jul 2013 | Pan MacMillan | 9781447236832 | English | London, United Kingdom Thierry Henry : Lonely at the Top PDF Book Bibliografische Informationen. Auclair is objective, almost to a fault, in his presentation and analysis. Associated Press. And yet, Henry has never shied away from the cameras. Readers also enjoyed. Part of HuffPost Sports. Worth a read to get a brief glimpse on the thought process of a great great footballer! This book is not yet featured on Listopia. Copyright Declaration. By continuing to use this website, you agree to their use. He was also instrumental in France's World Cup and European Cup glories in and and continued his incredibly successful career at Barcelona by winning the treble in his first season. I liked the book. But when I saw a book on Thierry Henry in a second hand book store at a throaway price, I could not resist. Philippe Auclair. But just like the Jets and the Giants, who play a little further north on the Turnpike, it's not quite New York. It goes on to talk about his early days in monaco, his first wc and his season in hell. We can safely leave the heavily promoted musings of Jessica Ennis, Bradley Wiggins, Sebastian Coe and Victoria Pendleton to fend for themselves, and turn instead to a couple of unauthorised life studies by writers whose experience, judgment and literary skills enable them to cast fresh light on their subjects. Rather, that seems to be the point of the book, until you reach the postscript, and a resolution.
    [Show full text]
  • Beckham Biography
    NEW EDITION OF ELLIS CASHMORE'S BECKHAM TO BE PUBLISHED ON 4 JUNE ‘Ellis Cashmore has compiled a remarkably stimulating and perceptive book on the phenomenon of Beckham. A significant contribution has been to place Beckham so acutely in the social context of today.’ JOHN GOODBODY, The Times ‘The second edition of this trail-blazing book reveals the secrets behind the “Beckham factor” and its intimate, symbiotic affair with the global media and the celebrity-driven twenty-first century. … Whether an icon or a cartoon, Beckham is different, and Cashmore knows like nobody else what buttons have been pushed to create the ultimate celebrity of a society where we are no longer citizens but consumers.’ MIGUEL ROZAS PASHLEY, El Mundo FULLY UPDATED FROM THE FIRST EDITION AND INCLUDING NEW FEATURES: • The pivotal role of Victoria in Beckham’s rise to iconic status; • A full exploration of the turbulent events of 2003 culminating in Beckham’s move to Real Madrid; • The scandal which hit the Beckham household in spring 2004. David Beckham remains as compelling as ever. As the complexity and contradictions of his character become more evident, so our enduring fascination with the world's premier sport celebrity becomes more puzzling. Why are we still spellbound by Beckham? The second edition of Ellis Cashmore's account of Beckham's life and times answers this question in ways that will provoke and animate. The key to understanding Beckham's undeniable power is not his playing ability, his good looks, his style, his family life or even his unerring ability to create headlines.
    [Show full text]
  • Candidates Short-Listed for UEFA Champions League Awards
    Media Release Route de Genève 46 Case postale Communiqué aux médias CH-1260 Nyon 2 Union des associations Tel. +41 22 994 45 59 européennes de football Medien-Mitteilung Fax +41 22 994 37 37 uefa.com [email protected] Date: 29/05/02 No. 88 - 2002 Candidates short-listed for UEFA Champions League awards Supporters invited to cast their votes via special polls on uefa.com official website With the curtain now down on the 2001/02 club competition season, UEFA’s Technical Study Group has produced short-lists of players and coaches in three different categories and will select a ‘Dream Team’ to mark the tenth anniversary of Europe’s top club competition, based on performances in the UEFA Champions League since it was officially launched in 1992. As usual, they have put forward nominees to receive the annual awards presented at the UEFA Gala, held as a curtain-raiser to the new club competition season at the end of August to coincide with the draws for the opening stages of the UEFA Champions League and the UEFA Cup. They have named six candidates for each of the awards presented to the Best Goalkeeper, Best Defender, Best Midfielder, Best Attacker, Most Valuable Player and Best Coach, based on their performances during the 2001/02 European campaign. This is not restricted to the UEFA Champions League – which explains the inclusion of Bert van Marwijk, head coach of the Feyenoord side which carried off the UEFA Cup and will be in Monaco to contest the UEFA Super Cup with Real Madrid CF.
    [Show full text]
  • Research Proposal
    DAVID BECKHAM’S EFFECT IN THE COVERAGE OF NEWS MEDIA A RESEARCH PAPER SUBMITTED TO THE GRADUATE SCHOOL IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE MASTERS OF ARTS BY ELIAS ESPINOSA DR. DUSTIN SUPA – ADVISOR BALL STATE UNIVERSITY MUNCIE, INDIANA NOVEMBER 2009 ii CONTENTS TABLES …………………………………………………………………………. Chapter 1. INTRODUCTION 2. LITERATURE REVIEW The beginning of Newspapers History of Soccer in the United States The Media, Culture and Sport Research Questions and Hypothesis 3. AMERICAN SOCCER BACKGROUND Situation analysis of soccer in the United States 4. METHODOLOGY ………………………………………………. 5. RESULTS ………………………………………………………… David Beckham in the L.A. Times Media’s focus of David Beckham 6. DISCUSSION …………………………………………………… Applications to Public Relations Implications for further research Conclusions iii Appendix 1. CODE SHEET ______________________________________________ 2. CODE BOOK ______________________________________________ REFERENCE LIST ______________________________________________ iv TABLES Table 1. Articles mentioning David Beckham before the announcement ______________ 2. Articles about David Beckham related to soccer or personal life _____________ v CHAPTER 1 INTRODUCTION The institutions of media and sports play an important role for the cross- promotion of each other industry. Together, look for mutual benefits when presenting information to their audiences. On one hand, the sport is promoted by the media coverage of the activities that the sport is involved in. On the other hand, the media is promoting them-selves by making its audience interested in the information being provided (Helland, 2007). It is clear that both institutions seek to gain attention from the public in general so their efforts can see the results in the revenues that they may obtain. Certainly, their motivation is the amount of money that is involved in these areas.
    [Show full text]
  • Learning to Rank with Joint Word-Image Embeddings
    Introduction Joint Word-Image Embedding Warp Loss Results Conclusion Web Scale Image Annotation: Learning to Rank with Joint Word-Image Embeddings Jason Weston∗, Samy Bengio∗ and Nicolas Usuniery ∗ Google, USA y Universite de Paris, LIP6, France BTAS, September 2010 1 / 28 Introduction Joint Word-Image Embedding Warp Loss Results Conclusion Image Annotation: What is it? Goal: Predict text given an image: 100,000s+ of possible annotations. ! obama ! eiffel tower What do we need? Scalable+good features ! we have not focused on this, use bag-of-terms Scalable(memory,speed)+good classifier ! One-Vs-Rest, PAMIR, k-NN? In this work we propose a scalable+good classifier. 2 / 28 Introduction Joint Word-Image Embedding Warp Loss Results Conclusion Datasets (to grasp the scale) Statistics ImageNet Web Number of Training Images 2,518,604 9,861,293 Number of Test Images 839,310 3,286,450 Number of Validation Images 837,612 3,287,280 Number of Labels 15,952 109,444 3 / 28 Introduction Joint Word-Image Embedding Warp Loss Results Conclusion Our Proposed Solution Model choice: Jointly embed annotations and images. Learns structure by multi-tasking (One-Vs-Rest doesn't) Small memory usage (82MB vs. 8.2GB for One-Vs-Rest) Fast Test Time (0.17s vs 0.5s for One-vs-Rest) (not including feature extraction time: 0.4s) New loss function: WARP Loss. Directly optimizes precision@k. Efficient to train on millions of examples. 4 / 28 Introduction Joint Word-Image Embedding Warp Loss Results Conclusion Our Proposed Solution 5 / 28 Introduction Joint Word-Image Embedding Warp Loss Results Conclusion Joint Word-Image Embedding Model Images: d = 10; 000 dimensional sparse \visterms".
    [Show full text]
  • Play Hard Sports Camp 380 S
    5th & 6th Grade (J) (continued) SPORTS BIOGRAPHIES (JB) Alexander Kwame. The Crossover JB Beckham. David Beckham Bruchac, J. Sports Shorts : An Anthology of Short JB Brady. Tom Brady Stories JB Bryant. Kobe Bryant Christopher, Matt. Ice Magic JB James . Lebron James Gutman, Dan. Shoeless Joe & Me JB Jordan. Michael Jordan Jeter, Derek. The Contract JB Kwan. Michelle Kwan Korman, Gordon. No More Dead Dogs JB Lee. Bruce Lee Lupica, Mike. Bat Boy Comeback Kids (series) JB Manning. Peyton Manning Game Changers (series) JB Pele . Pele Mack, W.C. Athlete Vs Mathlete (series) JB Robinson. Jackie Robinson Myers, Walter Dean. The Journal of Biddy Owens JB Rodgers. Aaron Rodgers Park, Linda Sue. Keeping Score JB Ruth. Babe Ruth Ritter, John. The Boy who saved Baseball Robinson, Sharon. Safe at Home Orange Public Library Children’s Services Orange Public Library & History Center Spinielli, Jerry. Crash 407 E. Chapman Ave., Orange CA 92866 Stone, Jeff. Tiger (714) 288 - 2400 Tocher, T. Chief Sunrise, John McGraw, and Me Taft Branch Library Tooke, Wes. Lucky : Maris, Mantle, and My Best 740 E. Taft Ave., Orange CA 92865 Summer Ever (714) 288 - 2430 El Modena Branch Library Wallace, Rich. Double Fake Play Hard Sports Camp 380 S. Hewes St., Orange, CA 92869 Wiles, Deborah. The Aurora County All-Stars (714) 288 - 2450 The Thrill of Victory & www.cityoforange.org/library/ Wolff, Virginia Euwer. Bat 6 the Agony of Defeat Yee, Lisa. Stanford Wong Flunks Big-time 7/15 1st & 2nd Grade (EI) Picture Books (E) 5th & 6th Grade (J) Carlson, Nancy. Louanne Pig in Making the Team Brown, Marc.
    [Show full text]
  • P18 Layout 1
    SPORTS THURSDAY, OCTOBER 31, 2013 Southampton soaring after brush with abyss LONDON: Four years after sliding into the English backed by former Southampton great Matt Le Cortese says he has no qualms about making diffi- now an England international, having scored with third tier and veering dangerously close to bank- Tissier failed with a bid to buy the club, but salva- cult decisions in the interests of the club. his first touch on his debut against Scotland in ruptcy, Southampton are riding high in fifth place tion arrived in the form of a Switzerland-based, “I think, to be honest, looking at those two sce- August. in the Premier League following a stunning trans- Germany-born entrepreneur called Markus narios, if one day I was in a similar situation again, I Central to Southampton’s success this season formation. Saturday’s 2-0 win at home to Fulham Liebherr. would do it the same way again,” Cortese told the has been an asphyxiating high-pressing game that meant that the modest club from the English south Liebherr had no previous experience of working Leaders in Football conference earlier this month. has yielded a 1-0 win at Liverpool and a 1-1 draw at coast have enjoyed the best start to a top-flight in football and promptly sacked head coach Mark “You have to make the decision, whether it is Manchester United, as well as a record of just three season in their 127-year existence. Wotte after buying the club, yet his arrival marked popular or unpopular.” With Cortese’s backing, goals conceded in nine league games.
    [Show full text]
  • Factsheet FIFA Awards
    FACT Sheet FIFA awards The FIFA World Player Gala has been in existence since 1991. In 2010, it was merged with France Football’s Ballon d’Or award to form the new “FIFA Ballon d’Or”. The newly named FIFA Ballon d’Or is a combination of the FIFA World Player of the Year award and France Football’s Ballon d’Or. The 2010 winner was chosen by national team coaches and captains as well as media representatives from all over the world. The FIFA Women’s World Player of the Year award was also decided by the votes of national team coaches and captains as well as media representatives from all over the world. For the first time, Coach of the Year awards were presented at the ceremony to honour the top coaching talent in both the men’s and women’s games. The voting panel had a similar make-up to those involved in selecting the winners of the FIFA Ballon d’Or and the FIFA Women’s World Player of the Year award. Awards presented: FIFA Ballon d’Or FIFA Women’s World Player of the Year FIFA World Coach of the Year for Men’s Football FIFA World Coach of the Year for Women’s Football FIFA Presidential Award FIFA Fair Play Award FIFA Puskás Award FIFA/FIFPro World XI FIFA Ballon d’Or / FIFA World Player of the Year Awards 2010 10.01.2011, Kongresshaus, Zurich men: 1. Lionel Messi (ARG) 2. Andrés Iniesta (ESP) 3. Xavi (ESP) women: 1. Marta (BRA) 2. Birgit Prinz (GER) 3.
    [Show full text]
  • Categories and Cliques
    Categories, Cliques and Analogies in Creative Information/Knowledge Management Tony Veale (+ students) School of Computer Science and informatics University College Dublin [email protected] http://Afflatus.UCD.ie 1 Cliques, Categories, Analogies, Metaphors and Ontologies Ontologies impose structure on a domain by grouping “like with like” Good categories have high intra-category similarity, low inter-category similarity The members of coherent (tight-knit) categories resemble social “cliques” Members closely connected: new members must bond with all existing members Analogy is a social connective: clique members have complementary roles Like socializes with Like: Consider bands, teams, summits, parties, matches, etc. Like socializes with Like à Like mingles with Like à Like collocates with Like ∴ Text Collocation à Category Co-Membership à Cliques as Categories 2 Categories are tight-knit clusters of instances in a Feature Space Furniture Creature Body_part Time Building E.g., Almuhareb & Poesio (2004/2005), Veale and Hao (2007/2008) 3 Manipulating Categories: The View from Cognitive Science / Ling. 4 Modelling Metaphor: The State of the Art in Computer Science Who wants a car that drinks gasoline? Fragile Systems / Tiny Coverage / Deep Knowledge Bottleneck / No Scalability 5 More Credible (& Useful) Goals: Lower Level, Broader Coverage A Modest Proposal Creative Information Mgmt. (texts, corpora, ngrams) Creative Knowledge Mgmt. (categories, ontologies) Creative information retrieval Metaphor finder / thesaurus Lexicalized Idea Exploration Robust Systems / Broad Coverage / Plentiful Shallow Knowledge / Scalability 6 The Supermarket Principle: Group by Theme and Kind Ontological Categories should reflect both semantic relatedness and similarity 7 Cliques: A Social-Model of Category Structure Brownie George Laura Scooter Dick 3-clique Karl Rummy 4-clique Judith Wolfie Condi Seymour Colin A Clique is a complete A clique is Maximal iff not sub-graph.
    [Show full text]
  • Argentina Diego Maradona
    Possible Favorite Male Players Goalkeepers Backline Midfielders Forwards Dida: Brazil Daniel Passarella: Argentina Diego Maradona: Argentina Herman Crespo: Argentina Rene Higuita: Columbia Carlos Alberto: Brazil Juan Sebastian Veron: Argentina Carlos Tevez: Argentina Petr Cech: Czech Republic Cafu: Brazil Roman Riquelme: Argentina Gabriel Batistuta: Argentina Gordan Banks: England Roberto Carlos: Brazil Roanldinho: Brazil Claudio Caniggia: Argentina Edwin Van der Sar: Holland Gilberto Silva: Brazil Kaka: Brazil Lionel Messi: Argentina Gianluigi Buffon: Italy Jhon Kennedy Hurtado: Columbia Diego: Brazil Pele: Brazil Dino Zoff: Italy John Terry: England Socrates: Brazil Robinho: Brazil Jorge Campos: Mexico Ashely Cole: England Carlos Valderrama: Columbia Ronaldo: Brazil Pereira Ricardo: Portugal Marcel Desailly: France Frank Ribery: France Rivaldo: Brazil Iker Casillas: Spain William Gallas: France Michel Platini: France Samuel Eto: Cameroon Andoni Zubizaretta: Spain Jaap Stam: Holland Zinedine Zidane: France Freddy Montero: Columbia Marcus Hahnemann: USA Ronald Koeman: Holland Eric Cantona: France Paul Scholes: England Tim Howard: USA Frank Rijkaard: Holland David Beckham: England Alan Shearer: England Chris Eylander: USA Frano Baresi: Italy Steven Gerrard: England Andrew Cole: England Kasey Keller: USA Paolo Maldini: Italy Ryan Giggs: England Wayne Rooney: England Gianluca Zambrotta: Italy Frank Lampard: England George Best: England Gennaro Gattuso: Italy Michael Ballack: Germany Michael Owen: England Emmanuel Bboue: Ivory Cloast Ruud
    [Show full text]
  • 25 Years of the Premier League
    Celebrating 25 years of the Premier League Introduction by John Fendley Getting Away With It Scarborough; nineteen eighty something. The ‘80’s kind of rolled into one for me until ‘89. For arguments sake let’s go with 82’. 8PM and I’m closing the kitchen door so I can listen to the second half of Liverpool’s latest foray into Europe. Back then you couldn’t watch. Listening was the only option; and for some reason you only got the second half. Don’t get me wrong, it was brilliant, I loved football on the radio, still do. If I’m lucky, the goals might be on Sportsnight. If not I’ll use my imagination. Point being - that football on TV back 03 in the greyties was in short shrift. Scarborough 1994. Monday Night and I’m round at my mate’s. For the record, the fact his dad has got Sky Sports is only a small contributing factor in our friendship. He’s a good lad. But back then “the dish” was a game changer. It wasn’t the norm. How many people do you know who’ve got an electric car? Thought not. Same question but change the car to a Sky Sports subscription – guarantee that in most homes not many hands would have gone up. If I’d had a set top box back then I would probably have added that information to my CV. Back to the football. Liverpool 3-3 United. This Monday Night Football could catch on. London 1996. Two years later and I am standing in the Sky Sports Canteen.
    [Show full text]
  • Controversial Changes MARCO VAN BASTEN's
    PROUD SPONSOR FREE COPY English & Spanish 01.27. 2017 | Year 1 | #18 | Miami, FL WS E FOR THOSE WHO Play, USN L THOSE WHO P WATCH AND ER THOSE WHO SUPPORT THE FUTURE OF SOCCER. US.US SOCC US.US L P ER US US WWW.SOCC L P ER SOCC A WS @US WS E Marco Van Basten’s USN L P ER SOCC CONTROVERSIAL YEAR 1 YEAR THE LEGENDCHANGES PROPOSES RULE MODIFICA TIONS TO REVOLUTIONIZE SOCCER >PAGES. 8-9 APRIL 30 / 2016 • JOGO BONITO THE CHAMPIONS CORONCORO SOCCER ATLETICO NACIONAL MIAMI REAL Madrid’s CUP IS READY TO GO CLUB IN MIAMI REOPENS ITS DOORS HERO PAGE 04 PAGE 11 PAGE 12 PAGE 14 2 PLANETSOCCERNEWS 01.27.2017 | Year 1 | #18 | Miami, FL THE R EDITORIAL FROM CCE O SOCCERPLUS PLANETNEWSS GOAL TOURNAMENTS SEASON ROAD TO BEIJING, 5V5 his weekend obtain the title in the Beijing 2017. The team throughout the tour- Alessandro Sbrizzo the excitement qualifier championship that wins in Miami will nament in Bangkok, Executive Director of 5v5 soc- for the COPA USA Na- advance to the finals F5WC Thailand 2016. [email protected] cer returns to tional 5V5. where they will decide The tournament is part TSouth Florida. The best who will be the USA rep- of the Football Fives teams in the area will Thirty two winning resentative. Colombia World Championship he present edition of Soccerplus be looking for a ticket to teams of the regional is the current reigning (FFWC). Don't miss out has spectacular topics including China Beijing 2017. They qualifier championships World Champ after be- this weekend at La Cai- Netherlands Legend Marco Van will also be seeking to will face off at F5WC ing crowned unbeaten manera in Miami.
    [Show full text]