Ink Features for Diagram Recognition - Rachel Patel, Beryl Plimmer, John Grundy, Ross Ihaka
Patel et al statistically determine which of the various features used in sketch recognition are the most useful for distinguishing shapes from text. They drew from a possible set of 46 features, including those previously used in sketch recognition and newly created features. First they collected sample sketches from 26 people for nine diagram types, each of which included shapes and text. From these samples, each feature was computed. Next, a complete statistical analysis was conducted to determine the most useful features. Using statistical partitioning, the most useful feature was chosen iteratively to create a decision tree. At each node in the tree, a binary test is performed based on the feature which will best decide on the two classes. For each outcome of the test, if the decision is not "good enough," the next best feature for that branch is chosen and another test is formed. Surprisingly, the tree is only 6 levels deep and employs only 8 features. When compared to the Microsoft Ink and InkKit dividers, the new method outperformed both method in Shape classification, but worse than the Microsoft in text classification (Ink seems to classify sketches leaning strongly towards text) on training data, and better on shapes on test data. On all set the text recognition was close to the recognition rate of the InkKit divider. Seemingly, the new recognizer show a considerable improvement in accuracy of division between text and shapes. The eight features used include: time till next stroke, Speed till next stroke, Distance from last stroke, Distance to next stroke, Bounding box width, Perimeter to area, Amount of ink inside, and Total angle.
Discussion - An overall accuracy rate would be useful in addition to the shape-only and text-only accuracies. The overall accuracy of the new method depends on how heavily the datasets were weighted towards shapes or text. If diagrams are shape heavy, the new recognizer obviously performs better, but if text is predominant, the others would be better. The distribution of text/shapes may be difficult a priori for many domains. Alternately, a combination recognizer perhaps using a voting scheme or boosting could be used to gain the advantages of the highly accurate text classifiers along with the accuracy of the shape recognizers.
skip to main |
skip to sidebar
Summary and discussion of assigned readings for the Fall 20007 Sketch Recognition class
About Me
The Required Stuff
Year: PhD 3rd year
Email: p r i s c u s 2 7 @ gmail.com
Academic Interests: Neural Networks, Human Cognition, Manifold Learning
Relevant Experience: Classes - Machine Learning, Pattern Classification
Why are you taking this class?
Not only does the subject sound interesting, it also seems like a technology that will be very relevant to the future use of computers. Also, it sounded like a cool class to gain some knowledge in the area of human-computer interaction in a hands-on way.
What do you hope to gain?
A better understanding of how humans use handwritten/drawn information to communicate and also how we understand this information. How people recognize what is depicted.
5 to 10 years from now?
Not really sure. Hopefully I'll have graduated. I'd like to go into a teaching career, preferably at the university level, though teaching at a lower level of education and being able to provide an introduction to computer science at the high school or younger level would also be interesting.
Nonacademic interests:
Swing (and other forms of) dancing, Computer and console gaming, Sci-fi and fantasy literature.
Fun Story:
Since this has recently been told to everyone I know, I'll tell it. When I was about 5 or 6 years old, we were heading out to the car to go to church. I was dressed to the nines, new shoes, new slacks, a button up shirt, hair slicked back. My mom looks over at me and says, "Well Paul, you sure look sharp today." And I instantly burst into tears. Because as we all know, one of the very first things you learn is that sharp is very, very bad and you should never touch sharp things.
Email: p r i s c u s 2 7 @ gmail.com
Academic Interests: Neural Networks, Human Cognition, Manifold Learning
Relevant Experience: Classes - Machine Learning, Pattern Classification
Why are you taking this class?
Not only does the subject sound interesting, it also seems like a technology that will be very relevant to the future use of computers. Also, it sounded like a cool class to gain some knowledge in the area of human-computer interaction in a hands-on way.
What do you hope to gain?
A better understanding of how humans use handwritten/drawn information to communicate and also how we understand this information. How people recognize what is depicted.
5 to 10 years from now?
Not really sure. Hopefully I'll have graduated. I'd like to go into a teaching career, preferably at the university level, though teaching at a lower level of education and being able to provide an introduction to computer science at the high school or younger level would also be interesting.
Nonacademic interests:
Swing (and other forms of) dancing, Computer and console gaming, Sci-fi and fantasy literature.
Fun Story:
Since this has recently been told to everyone I know, I'll tell it. When I was about 5 or 6 years old, we were heading out to the car to go to church. I was dressed to the nines, new shoes, new slacks, a button up shirt, hair slicked back. My mom looks over at me and says, "Well Paul, you sure look sharp today." And I instantly burst into tears. Because as we all know, one of the very first things you learn is that sharp is very, very bad and you should never touch sharp things.
No comments:
Post a Comment