skip to main
|
skip to sidebar
Recognition of Handwritten Names
Monday, February 23, 2009
Where the classifier is at now
Here is the ROC curve for the training data:
And for testing:
The curve is perfect for training but not excellent for testing. The next step is just to use what I have for my overall project.
0 comments:
Post a Comment
Newer Post
Older Post
Home
Subscribe to:
Post Comments (Atom)
Blog Archive
▼
2009
(14)
►
March
(5)
Presentation slides
Jittering Data
Well, I ran it already
Huh. How 'bout dat
Testing out the classifiers
▼
February
(5)
Where the classifier is at now
Improving the ROC curve
Boosting in Matlab
Boosting with OpenCV
Trying PCA
►
January
(4)
Fixing the cropping issue
Cropping
Fixing up training data
I'm baaaack
►
2008
(32)
►
March
(10)
Presentation slides
LoG filter not too hot either
Using a bigger binomial kernel
Incorporating another set of NN
Filtering with a discrete binomial kernel
Emission probabilities
Results after scaling
Looking into why my nearest neighbor results are a...
Viterbi getting a little better
Viterbi experiments
►
February
(9)
Experimenting with nearest neighbor
First try at nearest neighbor - very simple!
Test images all the same size
Example of test data
Training data all set
Looking at thresholded characters
Using results of edge.m
I'm better at getting rid of leftover lines!
Getting rid of leftover lines
►
January
(13)
Removing leftover edges
I got some training data
Example of cut out training data
Example of training data
Non maximal suppression not really working
Isolated letters
Better Isolated Boxes
Isolating character boxes
Better angles from Hough
Hough is working better but not there yet
Why is the Hough Transform not working?
Sample Input
Welcome to Dafna's CSE 190 blog!
0 comments:
Post a Comment