The Deep Learning course

2 Weeks

• Supervised Learning
• Unsupervised Learning

5 Weeks

• Logistic regression
• Neural Networks - CNN, RNN, LSTM
• Backpropagation, Deep networks
• Regularization, Dropout, Batch Normalization

5 Weeks

• Introduction
• Distributed word representations
• Language Modeling, Convolutional neural networks for Text
• GRUs, LSTMs for Language Modeling, Attention and Applications, GPT, BERTs and Variants
• Recurrent Neural Networks (Unidirectional and Bidirectional), Machine Translation, POS, Sentence Classification, Text Generation

4 Weeks

• Audio representations for deep learning
• Speech Recognition
• End-to-end deep networks
• Detection Models

5 Weeks

• Popular CNN architectures
• Transfer learning, autoencoders
• Object detection,image segmentation
• RNN and LSTM for image captioning/video

3 Weeks

• Introduction to sequential decision making under uncertainty
• Implementing RL algorithms with deep neural networks.
• Value functions, Finite and infinite Problems

4 Weeks

• Overview of various ML hardware for IoT/Edge devices
• Energy Efficiency, IoT/Edge Devices
• Optimization techniques * ML Model for Edge Devices

5 Weeks

• Deep Generative Models I
• Deep Generative Models II
• Semi and Self-supervised Learning I
• Semi and Self-supervised Learning II

• COVID-19 detection using acoustic sounds
• Use of NLP models for solving simple Math Word Problems
• Study of Practical BERT models for Sequence Labeling
• Prompt Learning for Vision Language Models
• Training a speech recogniser using data from speech synthesis
• Training a Voice conversion model
• Reinforcement Learning based Task Offloading in Mobile Edge Computing
• Anomaly Detection in Industrial IoT
• Federated Multi-Armed Bandits based Recommender System
• Interpretable GAN Controls
• Crafting Adversarial Samples for Deep Models
• Monocular Depth Estimation
• Text only training for Image Captioning
• Zero-shot segmentation using Stable Diffusion
• Open set zero shot classification via stable diffusion
• Knowledge distillation for Speech Recognition
• Time series forecasting via LLMs
• Lesion Boundary Segmentation
• Neural Models for Legal Judgement Prediction
• Optic-Disk-Cup Segmentation and Classification