CS 539 A99 - Project 2

Pose Recognizer. Train a neural network to recognize whether the person in a picture is looking up, straight, left, or right. For this, you just need to follow the steps of the training process described on Section 4.7 of the textbook. You should reproduce their experimental results.
Sunglasses Recognizer. Train a neural network to recognize whether or not the person in a picture is wearing sunglasses.

Code: You can use

the C code for the neural network error backpropagation algorithm from Chapter 4 of the textbook, or
your own code.
Your code must run on the CS or CCC Unix machines.
Training/Test Instances: Use the "quarter"-size collection of pictures provided at the CMU site. You can find copies of those pictures in our course public directory /cs/courses/cs539/s99/Projects/Proj2 on the CS machines.

Program and Neural Networks. You should submit (1) the source code of your program and (2) the neural networks that you obtained, by email to ruiz@cs.wpi.edu.
Written Report. Please bring your report to my office (FL232) or to class by the due date/time. Your report should discuss the following issues:
1. adaptation of the code (if any) or a description of your own code,
2. the experiments you ran with the system,
3. the topology (number of units in each hidden layer), initial weights, number of iterations of the error backpropagation algorithm, and final weights of each of your neural nets,
4. evaluation of the accuracy of each of your neural nets,
5. strengths and the weaknesses of your system.
Your report should also include a short user manual explaining how to install, run, and use your system (if different from the CMU package).
Oral Report. We will discuss the results from the individual projects during the class on March 2. Be ready to show your results (I strongly encourage you to prepare transparencies) and to discuss your project solution in class.