Deep Learning for Computer Vision

Week - 01

• Introduction to Convolutional Neural Network (CNN) architecture

• How to create a standard CNN architecture using TensorFlow

Week - 02

• Analyzing the application and effect of the following techniques in CNN - 

  • • • 1. Dropout Regularization

        2. Batch Normalization

        3. Batch Size

        4. Callbacks

Week - 03

• Applying a CNN architecture in a real-world dataset 

• Using a pre-trained model (VGG16) for prediction

• Data Augmentation

• How to build DL models in Kaggle

Week - 04

• Transfer Learning

  • • • • Pretrained model as feature extractor

        • Fine Tuning

Week - 05

• End-to-end DL/CNN pipeline for -

  • • • • Hand-drawn Electric circuit Schematic Components detection

        • Malaria parasite detection from blood smears

• How to build DL models in Google Colab

Week - 06

• Practical Example

  • • • • Skin cancer diagnosis - HAM10000 dataset

Week - 07

• Object Detection 

  • • • • R-CNN

        •  YOLO