Homework answers / question archive / Assignment 1 CPS584 - Advanced Intelligent Systems and Deep Learning Released Date: 09/09/2022 Requirements In this assignment, you will solve practical and interesting problems

Assignment 1 CPS584 - Advanced Intelligent Systems and Deep Learning Released Date: 09/09/2022 Requirements In this assignment, you will solve practical and interesting problems

Computer Science

Share With

Assignment 1
CPS584 - Advanced Intelligent Systems and Deep Learning
Released Date: 09/09/2022
Requirements
In this assignment, you will solve practical and interesting problems. By completing the project, you will gain valuable hands-on experience in the design, implementation and evaluation of classification algorithms. The details are listed as below.
You are provided with the “DogCat.zip” file which contains images of two classes: Dog and Cat. For each class, 25 training images and 15 testing images are given.
1. You need to resize (imresize function in MATLAB) the image to the size 48 x 48. Then, you write MATLAB code to extract the Haar-like features for each training and testing image. The details are as follows:
a. You compute the integral image (Lecture “Handcrafted Features”, slides 16-20).
2
b. By using the integral image, you compute the Haar-like features with 30 boxes below.
Note: use the integral image to compute the rectangle sum.
c. You then extract Haar-like features for both training and testing images. Each training/testing image has a 30-dim Haar-like feature.
2. You train 2 classifiers: K-nearest neighbor (KNN) and Neural Networks (NN) on the Haar-like features of training images. Then, you test the trained models on the Haar-like features of the testing images. Please report the accuracy rate for each class. You are requested to try different K values: 1, 3, 5, and 7.
3. Manually collect additional 30 training images for both classes: Dog and Cat. Note that those images must not be identical to any already given training/testing image. You may need to use any tool to resize and crop the newly collected images to the size 48 x 48.
4. Run 2 classifiers: K-nearest neighbor and Neural Networks on the testing images. Note that both classifiers (KNN and NN) are trained on the new training data – there are 25 (already provided) + 30 (newly collected) = 55 training images for each class. Again,
3
you are requested to try different K values: 1, 3, 5, and 7. Please report the accuracy rate for each class.
5. Discuss the accuracy rates in (2) and (4). Which one is better? Will more training data lead to a better performance?
What to Submit
1. A well-documented MATLAB program that implements the aforementioned problem in the Assignment 1. You must submit your program source code and the newly collected training data set.
2. A well-written, concise project report. It should include: (a) title and names of group members; (b) the analysis of each problem; (c) the issues during the implementation; (d) the solutions to overcome the issues in (c); (e) the contribution of each individual member; and (f) the powerpoint slides (maximum 20 slides) used in the Assignment presentation.
For each group, you must submit the files above in a single zipped folder. Your group will be required to do a face-to-face evaluation for the grading. One group member will submit the file on behalf of the group.
Important: Your submission will be thoroughly checked. If any plagiarism (from Internet, former students, or anywhere else) is found in this assignment, an F will be assigned to course grade and an academic dishonesty report will be given.
Submission Due: 11:55pm, October 5, 2022