Authors: Teaching Team
Target: Senior undergraduate students (Python programmers)
Duration: 16 sessions (8 effective weeks + Midterm + Final Exam)
Assessment Weight: 100% (see appendix)

Title Page

Deep Learning & Computer Vision
Project-Based One-Semester Textbook

Preface

This textbook is designed to meet the requirements of Deep Learning and Computer Vision courses within one semester (16 weeks). The materials progress from fundamental concepts to cloud and edge computing applications, using a project-based learning approach. Each chapter includes learning objectives, theoretical content, lab guides, and assessment instruments. The textbook also integrates modern platforms such as TensorFlow, PyTorch, YOLO, AWS/GCP, and Docker.

Table of Contents

• Part 1: Introduction and Deep Learning Fundamentals (Weeks 1–2)
• Part 2: Image Classification & Model Evaluation (Week 3)
• Part 3: Object Detection and Face Recognition (Weeks 4–5)
• Part 4: Cloud Computing, Deployment, and Edge AI (Weeks 6–7)
• Part 5: End-to-End Project and Case Studies (Week 8)

References & Appendices: Full Syllabus, Grading Rubric, Answer Keys

Week 1: Core Concepts of AI, Deep Learning, and Computer Vision

Session 1 – Theory: Introduction to AI, DL, CV

1.1 Definition of Artificial Intelligence (AI)

AI is the field of computer science focused on creating machines that can mimic human intelligence. AI is divided into Narrow AI (Weak AI) and General AI (Strong AI).

1.2 History of Deep Learning

1958 – Perceptron, 1986 – Backpropagation, 2012 – AlexNet, 2017 – Transformer.

1.3 Role of Computer Vision (CV)

Main CV tasks: image classification, object detection, semantic segmentation, image generation.

1.4 Industry Use Cases

Assignment: Initial quiz – 10 multiple-choice questions via LMS (Weight 5%)

Session 2 – Lab: Neural Networks and CNN

Perceptron, Activation Functions, Convolution, Pooling

import tensorflow as tf
from tensorflow.keras import layers, models
# NN for XOR gate
model = models.Sequential([
 layers.Dense(4, activation='relu', input_shape=(2,)),
 layers.Dense(1, activation='sigmoid')
])
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
X = [[0,0],[0,1],[1,0],[1,1]]
y = [0,1,1,0]
model.fit(X, y, epochs=200, verbose=0)
print(model.predict([[0,1]]))

Task: Coding exercise – build a NN for AND logic function (Weight 5%)

Session 3 – Lab: Image Preprocessing

Resize, normalization, augmentation, labeling. Augmentation example:

from tensorflow.keras.preprocessing.image import ImageDataGenerator
datagen = ImageDataGenerator(rotation_range=20, horizontal_flip=True)

Assignment: Structured dataset with 5x augmentation (Weight 5%)

Session 4 – Lab: Image Classification with CNN

CNN architecture: Conv2D + MaxPool + Flatten + Dense. Achieve at least 85% accuracy on custom dataset.

model = models.Sequential([
 layers.Conv2D(32, (3,3), activation='relu', input_shape=(32,32,3)),
 layers.MaxPooling2D((2,2)),
 layers.Conv2D(64, (3,3), activation='relu'),
 layers.MaxPooling2D((2,2)),
 layers.Flatten(),
 layers.Dense(64, activation='relu'),
 layers.Dense(10, activation='softmax')
])

Assignment weight: 5%

Week 3: Model Evaluation & Introduction to Object Detection

Confusion Matrix, Precision, Recall, F1-Score, ROC-AUC. Also YOLO, anchor boxes, IoU.

Weeks 4-5: Advanced Face Recognition & Object Detection

Content includes YOLO, SSD, and FaceNet embedding.

Week 6: Cloud AI and Deployment

Cloud services: AWS Rekognition, GCP Vision AI, Azure Computer Vision.

Week 8: End-to-End Pipeline & Final Presentations

Data pipeline, training pipeline, deployment pipeline. Case studies: Autonomous Vehicle, Surveillance, Retail Analytics.

Final Exam: Group project presentation and demo.