Feature request: Face-expression recommendation system

[PriyanshuLathi]

• Is your feature request related to a problem? Please describe:
  The current face expression recommendation system uses MobileNet, and there is a need to evaluate a custom CNN model built from scratch. Additionally, confusion matrix and ROC plots for model evaluation are not present.

• Describe the solution you'd like:
  Currently the accuracy using MobileNet is around 55% but I have implemented a custom CNN for face expression recommendation, which achieves an accuracy of 64.33%. I propose adding this model to the repository and generating AUC plots to evaluate its performance alongside MobileNet.

• Describe alternatives you've considered:
  The current MobileNet model is a solid choice, but using a custom CNN offers a different perspective and provides an alternative that could be useful for comparisons.

• Approach :

1. Integrate the custom CNN model and its performance evaluation.
2. Generate AUC plots for both the MobileNet and custom CNN models.
3. Compare and display results, including confusion matrix and ROC scores.

• Additional context:
  The custom CNN model is designed to offer an alternative to MobileNet, and the AUC plots will help evaluate and compare the performance of both models.

[github-actions]

Thanks for creating the issue in ML-Nexus!🎉
Before you start working on your PR,
Pull the latest changes to avoid any merge conflicts.

• Attach before & after screenshots in your PR for clarity.
• Include the issue number in your PR description for better tracking.
  Happy open-source contributing!☺️