Peekaboom: A Game for Locating Objects in Images

Slides:



Advertisements
Similar presentations
Write a Procedural Text
Advertisements

LABELING IMAGES LUIS VON AHN CARNEGIE MELLON UNIVERSITY.
English Paper One, Section B How many marks is Section B worth? 27 What is the percentage of your whole GCSE in English represented by those 27 marks?
How Do You Make a Program Wait?. 1. What is an algorithm? 2. Can you think of a reason why it might be inconvenient to program your robot to always go.
THE ESP GAME, & PEEKABOOM LUIS VON AHN CARNEGIE MELLON UNIVERSITY.
Extracting Valuable Information Lazily Shiry Ginosar Advisor: Professor Manuel Blum Graduate Mentor: Luis von Ahn.
1 Human Computation Play a Game to Develop an Ontology Peyman Nasirifard p+e+y+m+a+b-b+n dot deri.org.
Collaborative Human Computing Zack Zhu March 31, 2010 Seminar for Distributed Computing 1.
Crowdsourcing 04/11/2013 Neelima Chavali ECE 6504.
COVERT MULTI-PARTY COMPUTATION YINMENG ZHANG ALADDIN REU 2005 LUIS VON AHN MANUEL BLUM.
Lecture 26: Vision for the Internet CS6670: Computer Vision Noah Snavely.
Verbosities A game to collect common sense data Mihir Kedia.
Evaluating Hypotheses
CAPTCHA & THE ESP GAME SHAH JAYESH CS575SPRING 2008.
Need to know in order to do the normal dist problems How to calculate Z How to read a probability from the table, knowing Z **** how to convert table values.
Human Computation CSC4170 Web Intelligence and Social Computing Tutorial 7 Tutor: Tom Chao Zhou
Human Computation Steven Emory CS 575 Human Issues in Computing.
Advanced Technical Writing Lecture 8 Memorandums 29 June 2008.
Computer Engineering 294 R. Smith Outlines and Organization 10/ Organization, Outlines and Abstracts The objective of both written and verbal communication.
D ESIGNING G AMES WITH A P URPOSE By Luis von Ahn and Laura Dabbish.
© 2009 Mathew J. Sottile, Timothy G. Mattson, and Craig E Rasmussen 1 Concurrency in Programming Languages Matthew J. Sottile Timothy G. Mattson Craig.
Chapter 9 High-Level Programming Languages: C++. Chapter Goals Describe the expectations of high level languages Distinguish between functional design.
Rangers For our Rangers project we made a PowerPoint slide show highlighting math problems we created. First, we found out facts about the Rangers. Then,
Conditional & Joint Probability A brief digression back to joint probability: i.e. both events O and H occur Again, we can express joint probability in.
English 51 Wednesday, February 13, Free-write Today is “Get a Different Name” day. If you could change your name, would you, and why? What would.
Making Inferences and Drawing Conclusions
SCIENCE FAIR HINTS AND TIPS Charlotte Rodeen-Dickert St. Jerome School October 29,
Matthias Neubauer CAPTCHA What humans can do, But computers can not.
Mrs. Beth Cueni Carnegie Mellon
intelligence study and design of intelligent agentsis the intelligence of machines and the branch of computer science that aims to create it. AI textbooks.
Turing Test and other amusements. Read this! The Actual Article by Turing.
Partner Study Exciting Exponents!
Exploration Seminar 3 Human Computation Roy McElmurry.
Human Computation & ESP Game 2008/12/19 Presenter: Lin, Sin-Yan 1.
Lecture 5: Using Computers: Important Ideas Tonga Institute of Higher Education IT 141: Information Systems.
A problem is a doubtful or difficult question; a matter of inquiry, discussion, or thought; a question that exercises the mind (Oxford English Dictionary)
1-3. Answers will vary. Answers will vary, but robots can walk on water and upside down while only humans can feel emotions and write poems.
Computer Programming TCP1224 Chapter 3 Completing the Problem-Solving Process and Getting Started with C++
Labeling Images for FUN!!! Yan Cao, Chris Hinrichs.
Ideas Session Willer Travassos, Jan. 24th. GWAP Games with a purpose (GWAP) uses the computational power of humans to perform tasks that computers are.
Writing to convince others of your opinion..  Decide on your purpose: What will you convince the readers to believe or to do?  Pre-write to discover.
Advanced Technical Writing Lecture 4 Memorandums.
An Analytical Study of Puzzle Selection Strategies for the ESP Game Ling-Jyh Chen, Bo-Chun Wang, Kuan-Ta Chen Academia Sinica Irwin King, and Jimmy Lee.
Lesson 4 Using Variables in Python – Creating a Simple ChatBot Program.
Playing GWAP with strategies - using ESP as an example Wen-Yuan Zhu CSIE, NTNU.
CS315 Search and Retrieval of Multimedia P. Takis Metaxas.
CROWDSOURCING Massimo Poesio Part 2: Games with a Purpose.
Science Skills. Observations Observation - noticeable facts; using senses to gather information ; data measured, collected, perceived, or noticed 2 types:
1-3. Answers will vary. giraffe elephants warthog antelope.
ENGR-TS-2: The students will develop an understanding of how the design process is used to develop a technological system.
Welcome to MM570 Psychological Statistics
Video Games & Object Oriented Programming. Games.
Introduction to Professional Memo Writing
Human Computation (aka Crowdsourcing) LUIS VON AHN Slides taken from a talk by.
CS315 Multimedia Search and Retrieval. Types of Multimedia Multimedia: Beyond text communication Stored in a variety of formats Audio Today most popular.
Computational Thinking
Class Notes Unit 2 – Scientific Research Science:.
Unit 3 May I help you?. Teaching aims Do some exercise according to the text Read the passage loudly Understand the important role of a server Talk about.
Key Stage 2 SATs Information and Guidance on the Changes and Expectations for 2015/16 Mathematics Katy Piazza.
CS 2750: Machine Learning Active Learning and Crowdsourcing
1 Quiz Show Programming Terms. 2 Alice - 3D Virtual Programming Vocabulary Quiz Board Chapter 1 Chapter 2a Chapter 2b Chapter 3 Chapter 4 $100 $200 $300.
THE ESP GAME, AND OTHER STUFF
Inferences Fact vs. Opinion.
Max LaRue,Samrach Nouv, Yih Sun Khoo
Mrs. Beth Cueni Carnegie Mellon
A game to collect common sense data
Key Stage 1 National Curriculum
Key Stage 2 SATs Earsham Primary School Presentation to Parents
Tonga Institute of Higher Education IT 141: Information Systems
Presentation transcript:

Peekaboom: A Game for Locating Objects in Images Roy Liu Carnegie Mellon University Joint work with Luis von Ahn and Manuel Blum

Object Location in Images Given an image, determine what objects there are and locate them: Woman Man Umbrella Tree Sailboat Dog

As Things Stand Now No algorithm is known for taking an image and determining what objects are in it, let alone telling you where they are. Fortunately, this talk isn’t about developing such an algorithm. Let’s try a different approach.

Harnessing the Power of Human Cycles “Math is hard. Let’s go shopping!” –Barbie On similar line of thinking: Programming computers to locate objects in images is hard, so… Let’s not think about it. Instead, let’s get humans to do the work for us.

Problems Wait! Your average human probably wants: Enjoyment – they want to have a good time Incentives – they want something in return How do we address both?

A Game Have people do the work for us by playing a game. Many design issues arise: What will be the core idea of the game? How do we collect data? How do we ensure the quality of the data?

An Earlier Idea: Luis von Ahn’s ESP Game Two players, with no communication, each try to guess what the other is thinking about a particular image they both see. If they agree on a word, the game moves on and increases both players’ scores.

Server: Agreed, “Model” A Sample Run Player 1 Guesses Pants Model Lady Player 2 Guesses Woman Shirt Denim Girl Model Server: Agreed, “Model”

Why ESP Works By agreeing on a word, the players: Say what it is – we call this assigning a “label” to the image. Check their own work – the fact that two strangers agree on a label is a witness of the label’s quality.

The Limitations of ESP The ESP Game can label images (and consequently tell you what’s in them), but it cannot: Find the objects being labeled. Determine the way in which the object appears – does the label “car” refer to the text “car” or an actual car in the image?

Completing the Image Cycle ESP game server unlabeled images labeled images Peekaboom game server located images

A New Idea: Peekaboom Two players are assigned the roles of “revealer” and “guesser”. The revealer sees an image with a label. The guesser sees nothing. The revealer shows the guesser parts of the image. If the guesser guesses correctly, the game moves on.

Statement of Purpose We would like to collect data about images systematically and en masse. We hope our collection will form the basis for data sets that can be used to train computer vision algorithms.

A (Simplified) Trial Run

The Guesser guesses: Flower Petal Butterfly The Revealer clicks on parts of the image and shows them to the Guesser. Server: Correct, Butterfly

Why Peekaboom Works By getting the guesser to guess correctly, the revealer locates objects by clicking on the relevant parts of the image:

But Wait, There’s More Peekaboom not only locates objects, but also: Gives the context necessary to identify them. Classifies the image as “Text”, “Noun”, or “Verb” by way of hints. Let’s examine how Peekaboom does both.

Object Context The label: trunk Pings help separate the context of object with the object itself. They help the guesser distinguish trunk from other possibly correct labels like “elephant”, “tusk”, and “ear”.

Hints The hints help distinguish the manner in which the label “car” appears: this is the object “car” this is the text “car” The label “car” is ambiguous -- this is “car” this is also “car”

The Architecture of Peekaboom compiler labeled images players raw data game server located images players researchers

Peekaboom is… fun novel aesthetically appealing networked scalable widely deployable

Conclusions Peekaboom will be released to a general audience within two months. We hope that it will solve difficult AI tasks while achieving: Low costs – One game server. Quality – Accurately locate objects in images. Quantity – Locate objects in millions of images.

Advertisement The best way to understand this talk is to try the game out for yourself: www.peekaboom.org We look forward to collecting your cycles!