🧠 Deep Learning & Computer Vision

🕒 Duration: 12 Weeks
🎯 Prerequisite: Foundations of AI & Machine Learning Course
🛠️ Tools & Tech:

• Python

• Jupyter Notebooks / Google Colab

• TensorFlow (with Keras) – Core deep learning framework

• NumPy – Numerical computation

• Matplotlib & Seaborn – Visualization

• OpenCV – Image processing

• Joblib / Pickle – Saving models

Internship: 1 Month Internship program for selected candidates @ Srivin Tech Private Limited

📚 Topics Covered:

1. Introduction to Deep Learning

   • What is Deep Learning?

   • Difference between ML & DL

   • Overview of Neural Networks

2. Python Refresher for DL

   • NumPy basics (tensors, matrices)

   • Plotting learning curves with Matplotlib

3. TensorFlow & Keras Basics

   • Tensors, operations, and model architecture

   • Building a neural network using Keras Sequential API

   • Activation functions: ReLU, Sigmoid, Softmax

4. Training Deep Learning Models

   • Forward & backward propagation

   • Loss functions and optimizers (MSE, Cross-Entropy, Adam)

   • Overfitting, underfitting, and regularization techniques

5. Convolutional Neural Networks (CNNs)

   • Introduction to CNN architecture

   • Filters, strides, pooling layers

   • Building image classifiers using CNNs

6. Image Processing with OpenCV

   • Image loading, resizing, grayscale, thresholding

   • Edge detection and noise reduction

   • Using OpenCV images as input for CNN models

7. Model Evaluation & Tuning

   • Accuracy, confusion matrix, precision, recall

   • Plotting loss & accuracy curves

   • Using validation & test sets, dropout layers, early stopping

8. Model Saving & Reuse

   • Save and load models using Joblib or model.save() in Keras

   • Export models for future predictions

9. Mini Projects

   • Project 1: Handwritten Digit Recognition (MNIST)

   • Project 2: Face Detection / Mask Detection with OpenCV + CNN

   • Optional: Object classification using a real-world image dataset

✅ Outcome:

By the end of this course, learners will be able to design, train, and evaluate deep neural networks, work with image datasets, and deploy real-world computer vision applications using TensorFlow and OpenCV.