Deep-Learning-Course

Welcome to the Deep Learning course repository offered at the University of Tehran. This repository contains code for assignments and projects completed during the course. The course by:

• Dr. Ahmad Kalhor

Course Description

This course offers a comprehensive exploration of foundational principles in Deep Learning. It begins with a focus on basic concepts, starting from single-layer neural networks, gradually progressing through Convolutional Neural Networks (CNNs), Faster R-CNNs, and Recurrent Neural Networks (RNNs) including Long Short-Term Memory (LSTM). Continuing forward, it delves into advanced architectures such as Transformers and culminates with Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs).

Table of Contents

Please find below a brief overview of the contents of this repository:

1. HW1/:

   • Madeline and Adeline: Implement simple perceptron and Adaline networks for the Iris dataset.
   • DAC: Deep Autoencoder-based Clustering: Implementation of DAC, a general deep learning framework for representation learning.
   • Knowledge Distillation on MNIST: Implementation of knowledge distillation applied to the MNIST dataset.

2. HW2/:

   • Implementation of CNN-based Facial Affect Analysis on Mobile Devices.
   • Implementation of An Efficient CNN Model for COVID-19 Disease Detection Based on X-Ray Image Classification.

3. HW3/:

   • Fine-tuning segment anything model.
   • Implementation of faster RCNN from scratch.

4. HW4/:

   • LSTM for Time series data.
   • Implementation of Stacked CNN - LSTM approach for prediction of suicidal ideation on social media.

5. HW5/:

   • Fine-tuning HuBERT for emotion recognition on ShEMO: Sharif Emotional Speech Database.
   • Fine-tuning BERT.

6. HW6/:

   • Implementation of Controlvae: Controllable variational autoencoder.
   • Implementation of GANs: Wasserstein GAN and Self-Supervised GAN.

7. HWEXTERA/:

   • Fine-tuning using LoRA on the MultiNLI dataset.
   • Implementation of A Convolutional Neural Network Model for Credit Card Fraud Detection.

Disclaimer

This repository is for archival and reference purposes only. The code here might not be updated or maintained. Use it at your own discretion.