Project 4 - Face Detection

Brian Le
CSE 455
Winter 09

I could not finish the program with enough time to complete the write-up.
My main problem was testing too so long that to fix small bugs would take 15 minutes each at least.
My suggestion is to have the code due early, then the artifact due a day or two later like we had for project 1.
I do accept that it was my fault for not planning ahead that the artifact would take time

My program at this point I believe is all correct except for the findFace method
that we had to implement. It does not seem to be able to correct identify a face
which maybe a problem with the other methods I wrote in the eigfaces.cpp. However,
I don't see what was buggy with that code.

When I try to catch for overlap issues, my program's runtime significantly lengthens.
Therefore, I commented it out for the binary. I also commented out the use of steps and scaling.
As a last effort, I messed up my code for find face so the list is never updated with a better face.
Why, I'm not sure.

Testing recognition with cropped class images

Average Face

Eigenfaces

1. Describe the trends you see in your plots. Discuss the tradeoffs; how many eigenfaces should one use? Is there a clear answer?
2. You likely saw some recognition errors in step 3; show images of a couple. How reasonable were the mistakes? Did the correct answer at least appear highly in the sorted results?

Cropping and Finding Faces

1. What min_scale, max_scale, and scale step did you use for each image?
2. Did your attempt to find faces result in any false positives and/or false negatives? Discuss each mistake, and why you think they might have occurred.

Testing recognition with cropped class images

1. What MSE thresholds did you try? Which one worked best? What search method did you use to find it?
2. Using the best MSE threshold, what was the false negative rate? What was the false positive rate?