Skip to main content
B.Eng DegreeFaculty of Faculty of Engineering & TechnologyComputer Engineering

CPE 403: Advanced Embedded Systems & Edge AI

TinyML, model quantization (int8), pruning, TensorFlow Lite Micro, neural accelerators (NPU), and low-power inference.

Instructor
Created by PAMCET Learning Team
advanced2 - 4 weeksEnglish

Course Overview

TinyML, model quantization (int8), pruning, TensorFlow Lite Micro, neural accelerators (NPU), and low-power inference. Curriculum Structure & Core Modules: • 1. The TinyML Paradigm: Edge Artificial Intelligence Constraints, Latency, Privacy, and Energy Budgets • 2. Model...

What You'll Learn

  • Train, quantize (float32 to int8), and prune deep learning models for deployment on constrained microcontrollers.
  • Deploy Convolutional Neural Networks (CNNs) using TensorFlow Lite for Microcontrollers (TFLM) and Edge Impulse.
  • Optimize memory footprint to execute inference within restricted MCU SRAM (<256 KB) and Flash limits.
  • Evaluate dedicated edge AI neural processing unit (NPU) accelerators and DSP instruction extensions (CMSIS-NN).
  • Design real-time audio keyword spotting and vision anomaly detection systems on edge hardware.

Requirements

  • Completion of CPE 302.

Degree Programme Curriculum Context

This course is a curriculum unit of the Bachelor of Engineering (B.Eng.) in Computer Engineering degree framework in Nigeria.

Earn Your Certificate

Showcase your skills with a certificate! Complete the course, submit your project, and earn your certificate. Here's a sample of what you'll receive to show off your accomplishments.

course-certificate
Lecture Board & Syllabus

Course Curriculum

Browse complete theoretical foundations, worked calculations, laboratory protocols, and board assessment questions under each topic.

Launch Course Reader →
8 Chapters
•
16 Lecture
•
24h
Subtopic 1Document LectureOpen Access~3 min read

The TinyML Paradigm: Edge Artificial Intelligence Constraints, Latency, Privacy, and Energy Budgets: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 1 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

The TinyML Paradigm: Edge Artificial Intelligence Constraints, Latency, Privacy, and Energy Budgets: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 1 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 1 Formative Assessment: The TinyML Paradigm: Edge Artificial Intelligence Constraint

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 1: The TinyML Paradigm: Edge Artificial Intelligence Constraints, Latency, Privacy, and Energy Budgets. Aligned with NUC CCMAS and COREN OBE standards.

Module 1 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

Model Optimization Pipeline: Post-Training Quantization, Quantization-Aware Training (QAT), and Weight Pruning: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 2 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

Model Optimization Pipeline: Post-Training Quantization, Quantization-Aware Training (QAT), and Weight Pruning: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 2 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 2 Formative Assessment: Model Optimization Pipeline: Post-Training Quantization, Qua

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 2: Model Optimization Pipeline: Post-Training Quantization, Quantization-Aware Training (QAT), and Weight Pruning. Aligned with NUC CCMAS and COREN OBE standards.

Module 2 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

TensorFlow Lite for Microcontrollers (TFLM): FlatBuffers, Memory Arenas, and Tensor Allocators: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 3 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

TensorFlow Lite for Microcontrollers (TFLM): FlatBuffers, Memory Arenas, and Tensor Allocators: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 3 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 3 Formative Assessment: TensorFlow Lite for Microcontrollers (TFLM): FlatBuffers, Me

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 3: TensorFlow Lite for Microcontrollers (TFLM): FlatBuffers, Memory Arenas, and Tensor Allocators. Aligned with NUC CCMAS and COREN OBE standards.

Module 3 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

CMSIS-NN: ARM Cortex-M DSP Vector Optimization for Quantized Neural Network Convolutions: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 4 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

CMSIS-NN: ARM Cortex-M DSP Vector Optimization for Quantized Neural Network Convolutions: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 4 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 4 Formative Assessment: CMSIS-NN: ARM Cortex-M DSP Vector Optimization for Quantized

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 4: CMSIS-NN: ARM Cortex-M DSP Vector Optimization for Quantized Neural Network Convolutions. Aligned with NUC CCMAS and COREN OBE standards.

Module 4 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

Audio Keyword Spotting on Edge Hardware: Mel-Frequency Cepstral Coefficients (MFCC) Extraction on MCUs: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 5 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

Audio Keyword Spotting on Edge Hardware: Mel-Frequency Cepstral Coefficients (MFCC) Extraction on MCUs: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 5 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 5 Formative Assessment: Audio Keyword Spotting on Edge Hardware: Mel-Frequency Cepst

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 5: Audio Keyword Spotting on Edge Hardware: Mel-Frequency Cepstral Coefficients (MFCC) Extraction on MCUs. Aligned with NUC CCMAS and COREN OBE standards.

Module 5 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

Computer Vision at the Edge: Person Detection with MobileNet and Micro-Image Classifiers: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 6 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

Computer Vision at the Edge: Person Detection with MobileNet and Micro-Image Classifiers: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 6 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 6 Formative Assessment: Computer Vision at the Edge: Person Detection with MobileNet

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 6: Computer Vision at the Edge: Person Detection with MobileNet and Micro-Image Classifiers. Aligned with NUC CCMAS and COREN OBE standards.

Module 6 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

Edge AI Accelerators: Specialized Neural Processing Units (NPUs), Google Coral, and Kneron Architectures: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 7 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

Edge AI Accelerators: Specialized Neural Processing Units (NPUs), Google Coral, and Kneron Architectures: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 7 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 7 Formative Assessment: Edge AI Accelerators: Specialized Neural Processing Units (N

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 7: Edge AI Accelerators: Specialized Neural Processing Units (NPUs), Google Coral, and Kneron Architectures. Aligned with NUC CCMAS and COREN OBE standards.

Module 7 • Formative Assessment
Subtopic 1Document LectureOpen Access~3 min read

Lifecycle and Reliability: Drifting Sensors, On-Device Learning, and Energy-Harvesting Operation: Theoretical Foundations, Modeling & Architecture

CPE 403 &bull; 400L &b

Module 8 • Subtopic 1•500 characters of authentic academic content
Subtopic 2Document LectureOpen Access~3 min read

Lifecycle and Reliability: Drifting Sensors, On-Device Learning, and Energy-Harvesting Operation: Computations, Implementation & Verification

CPE 403 &bull; 400L &b

Module 8 • Subtopic 2•500 characters of authentic academic content
Formative Assessment10 Points • 15 Mins

Chapter 8 Formative Assessment: Lifecycle and Reliability: Drifting Sensors, On-Device Learn

Assessment Protocol:10 Total Points•70% Passing Threshold•15 Mins Time Limit•Bloom's Taxonomy: Analyze & Evaluate

Rigorous formative competency assessment for Chapter 8: Lifecycle and Reliability: Drifting Sensors, On-Device Learning, and Energy-Harvesting Operation. Aligned with NUC CCMAS and COREN OBE standards.

Module 8 • Formative Assessment

Instructor

PAMCET Learning Team

PAMCET Learning Team

Institutional Course Curation · PAMCET

The PAMCET Learning Team is the institutional curriculum and content group behind course publishing on this platform. PAMCET (Professional and Academic Management Centre for Education and Training) is operated by Digitpen Hub Ltd, a company registered in Lagos, Nigeria. Courses credited to the PAMCET Learning Team are structured, organized, and published centrally by the platform rather than authored by a single named individual instructor — each course's syllabus, learning outcomes, and content are reviewed before publication under this institutional account. For questions about a specific course or to report a content issue, please use the platform's Help & Support or Contact Us pages.

Qualifications:

PAMCET Learning Team — Institutional Course Curation, Digitpen Hub Ltd

My Skills:

  • Curriculum Curation
  • Course Content Review
  • Learning Outcome Design
  • Course Publishing & Quality Assurance
  • Platform Content Operations

Reviews

No reviews available
CPE 403: Advanced Embedded Systems & Edge AI
Free$24.99100% Off
Certificate available separately for $7 after completion

Course Includes

8 Chapters
16 Lectures
2 - 4 weeks
Skill Level advanced
Taught in English
Lifetime Access
Certification of Completion
Category: University - Computer Engineering
Share On

Download Syllabus & Study Guide

Receive the full lecture syllabus, reading list, and milestone schedule for CPE 403: Advanced Embedded Systems & Edge AI.

By submitting, you agree to receive educational updates. You can unsubscribe anytime. Read our Privacy Policy.

CPE 403: Advanced Embedded Systems & Edge AI | PAMCET University — B.Eng | PAMCET