Course Sessions
Comprehensive modules covering computer vision from fundamentals to deployment
Introduction to Computer Vision
Fundamentals, applications, CV tasks (classification, detection, segmentation), and development environment setup
Session 2Data Preparation & Exploration
Data sources, annotation tools, quality analysis, class imbalance handling, and data organization
Session 3Feature Engineering & Preprocessing
Image transformations, data augmentation with Albumentations, classical features (SIFT, HOG), and dimensionality reduction
Session 4CNN Architectures & Modeling
CNN fundamentals, VGG, ResNet, EfficientNet, Vision Transformers, YOLO, U-Net, and training configuration
Session 5Transfer Learning & Optimization
Transfer learning strategies, performance evaluation, model diagnosis, regularization, and hyperparameter tuning
Session 6Production Deployment
Model optimization, ONNX, TensorRT, FastAPI deployment, Docker, MLOps with MLflow, and edge deployment
Practical Works
Hands-on exercises to reinforce each session's concepts
Getting Started
Environment setup, image loading with OpenCV, MNIST exploration and visualization
Lab 2Dataset Analysis
Real-world dataset exploration, statistical analysis, quality checks, and train/val/test splits
Lab 3Preprocessing Pipeline
Complete augmentation pipeline with Albumentations, PyTorch DataLoader, and visualization
Lab 4Training from Scratch
Implement CNN with Keras/PyTorch, train on CIFAR-10, visualize training curves
Lab 5Transfer Learning in Practice
Fine-tune EfficientNet, compare strategies, experiment tracking with W&B
Lab 6Model Deployment
Build FastAPI inference service, containerize with Docker, deploy to cloud