Completion of Chest X-Ray Images (Pneumonia) Detection Demo Using Ivy Framework

[leebissessar5]

Changes Proposed in this Pull Request:

• Pneumonia Detection with Ivy: This notebook demonstrates pneumonia detection in chest X-ray images using the Ivy machine learning framework.
• Data Preprocessing with PyTorch: The preprocessing of images is handled by PyTorch's built-in transformation modules, illustrating Ivy's interoperability with other libraries.
• Model Definition with Ivy.Module: The classifier is built using Ivy's Module class, providing a clear and structured way to define the model's architecture.
• Training and Validation: These are executed with an informative progress bar to monitor the experiment's advancement.
• Evaluation: A classification report and a confusion matrix are used to quantify the model's performance.
• Pretrained Weights: Weights for a pretrained model (achieving ~80% test accuracy) are provided.

Impact of Change:

• Ivy Demonstration: This notebook serves as a practical example of using Ivy for medical image classification, highlighting its capabilities and ease of use.
• Flexibility: The integration with PyTorch modules underscores Ivy's versatility and its ability to work with familiar tools in the deep learning ecosystem.

How to Test:

1. Setup: Follow the instructions in the notebook to install dependencies and set up the environment.
2. Run: Execute the notebook cells sequentially to train the model, evaluate its performance, and visualize results.

Additional Notes:

• Dataset: The dataset, sourced from Kaggle, exhibits a class imbalance skewed towards pneumonia cases. Recommendations for addressing this are provided within the notebook, including techniques like oversampling and more advanced architectures.
• Model Size and Training Speed: A deliberately small model was chosen to:
  • Accommodate the computational resources of Colab's free tier.
  • Enable faster training times.
  • Allow for experimentation on less powerful GPUs.
  • Emphasize the primary goal of showcasing Ivy's functionality.

Closes #91