#Project Overview

This project is done for the assignment for the 7154CEM - Artificial Intelligence module at Coventry University.

The goal of this project is to design, impleament and compare two convulutional nueral networks to classify the images from CIFAR-10 dataset.

This repository contains two parts: - Part A: Human designed and impleamented model to classify the images from the CIFAR - 10 dataset. - Part B: Generative AI assisted model, designed with suggestions from the Generative AI tools (Claude), where the prompts and responses are recorded.

Both the models are evaluated using: - Accuracy (%) - Confusion matrices - Misclassified images detection - Visual inspection of model uncertainty

A critical comparision and reflection is included in the final report and short presentation video.

#Steps to run this project

Step 1: Clone the repository using the below command:

-- git clone https://github.coventry.ac.uk/pawatec/7154-ANN_Coursework.git
-- cd 7154-ANN_Coursework

Step 2: Setup the Python Environment

-- pip install torch torchvision matplotlib numpy seaborn scikit-learn

Step 3: To Run Part A -- cd Part_A -- python PartA.py

Step 4: To Run Part B -- cd Part_B -- python PartB_Model_Evaluation.py

How to Run the Notebooks in Google Colab The recommended way to run these notebooks is using Google Colab, as it provides free access to GPUs and requires no local setup.

Step 1: Download the Project Files First, you need to download the project files from this GitHub repository to your local computer.

1. On the main page of this repository, click the green <> Code button.
2. Select Download ZIP.
3. Save the ZIP file to your computer and then unzip it. You will now have a folder containing all the project files.

Step 2: Upload and Open the Notebook in Colab Next, you will upload the notebook you want to run into your Google Colab environment.

1. Go to Google Colab.
2. In the pop-up window, click on the Upload tab.
3. Click "Choose File" and navigate to the unzipped project folder you downloaded.
4. Select the .ipynb notebook file you wish to run.
    -- For the Human-Guided model, it will be inside the Human-Guided Neural Network folder.
    -- For the AI-Guided models, they will be inside the GEN AI Guided Neural Networks folder.

Step 3: Enable GPU Acceleration For training neural networks, using a GPU is essential for speed. Before running any code, make sure to enable the GPU accelerator.

1. Once the notebook is open, click on Runtime in the top menu.
2. Select Change runtime type.
3. In the pop-up window, select GPU from the "Hardware accelerator" dropdown menu.
4. Click Save.

Step 4: Run the Code Once the notebook is open and the GPU is enabled, you can run the code. -- To run all cells sequentially:** Go to Runtime -> Run all. -- To run cells one by one:** Click on a cell and press Shift + Enter.

pawatec/7154-ANN_Coursework

About

Resources

Stars

Watchers

Forks

Releases

Languages