Module Details

Module Code: COMP C9025
Full Title: Computer Vision and AI
Valid From:: Semester 1 - 2019/20 ( June 2019 )
Language of Instruction:English
Duration: 1 Semester
Credits:: 7.5
Module Owner:: John Loane
Departments: Unknown
Module Description: Learn to use Deep Learning, Computer Vision and Machine Learning techniques to build autonomous cars, robots, and drones that can see, respond and learn from their environments.
 
Module Learning Outcome
On successful completion of this module the learner will be able to:
# Module Learning Outcome Description
MLO1 Apply computer vision techniques to identify objects
MLO2 Design and apply a perceptron-based neural network to classify between binary classes
MLO3 Design and apply convolutional neural networks that can classify images
MLO4 Design and apply deep neural networks to fit complex datasets
MLO5 Build and train a robot, drone or car that can operate on their own
Pre-requisite learning
Module Recommendations
This is prior learning (or a practical skill) that is strongly recommended before enrolment in this module. You may enrol in this module if you have not acquired the recommended learning but you will have considerable difficulty in passing (i.e. achieving the learning outcomes of) the module. While the prior learning is expressed as named DkIT module(s) it also allows for learning (in another module or modules) which is equivalent to the learning specified in the named module(s).
No recommendations listed
 
Module Indicative Content
Computer Vision
We will study how to use open source computer vision to detect objects in images, edges, regions of interest and find lanes on video.
Perceptrons
Use perceptrons to classify a linearly separable set of inputs.
Deep Neural Networks
Will will study deep neural networks to help us extract features from images.
Multiclass Classification
We will study how to train a model to fit more than just two classes. We will use this later to classify images.
Image recognition
We will use openly available image datasets to see how to train, validate and test image recognition.
Convolutional Nueral Networks
Convolutional neural networks are the go-to model for image classification. We will study these efficient way to extract features from images.
Classifying Images
We will study how to process data to train a neural network to classify road signs.
Polynomial regression
We will learn how to build a model to fit a continuous dataset.
Behavioural Cloning
We will use the convolutional neural network to learn from training data.
Building Robots
We will cover how to build a robot with simple sensors, cheap computers and open source tools.
Module Assessment
Assessment Breakdown%
Course Work100.00%
Module Special Regulation
 

Assessments

Full Time On Campus

Course Work
Assessment Type Continuous Assessment % of Total Mark 50
Marks Out Of 0 Pass Mark 0
Timing Week 9 Learning Outcome 1,4,5
Duration in minutes 0
Assessment Description
Students will develop a robot or drone which can recognise objects.
Assessment Type Continuous Assessment % of Total Mark 50
Marks Out Of 0 Pass Mark 0
Timing Week 13 Learning Outcome 1,2,3,4,5
Duration in minutes 0
Assessment Description
Students will develop a simulation of a self-driving car which can identify lanes and traffic signs.
No Project
No Practical
No Final Examination

Part Time On Campus

Course Work
Assessment Type Continuous Assessment % of Total Mark 50
Marks Out Of 0 Pass Mark 0
Timing Week 9 Learning Outcome 1,4,5
Duration in minutes 0
Assessment Description
Students will develop a robot or drone which can recognise objects.
Assessment Type Continuous Assessment % of Total Mark 50
Marks Out Of 0 Pass Mark 0
Timing Week 13 Learning Outcome 1,2,3,4,5
Duration in minutes 0
Assessment Description
Students will develop a simulation of a self-driving car which can identify lanes and traffic signs.
No Project
No Practical
No Final Examination
Reassessment Requirement
No repeat examination
Reassessment of this module will be offered solely on the basis of coursework and a repeat examination will not be offered.

DKIT reserves the right to alter the nature and timings of assessment

 

Module Workload

Workload: Full Time On Campus
Workload Type Contact Type Workload Description Frequency Average Weekly Learner Workload Hours
Lecturer Supervised Learning Contact The delivery of theory will be integrated with the practical implementation of that theory. Classes will normally be broken down into 40% theory delivery and 60% practical implementation. Every Week 3.00 3
Directed Reading Non Contact Reading of lecturer-recommended information sources. Every Week 3.00 3
Independent Study Non Contact Independent practical work. Every Week 6.00 6
Total Weekly Learner Workload 12.00
Total Weekly Contact Hours 3.00
Workload: Part Time On Campus
Workload Type Contact Type Workload Description Frequency Average Weekly Learner Workload Hours
Lecturer Supervised Learning Contact The delivery of theory will be integrated with the practical implementation of that theory. Classes will normally be broken down into 40% theory delivery and 60% practical implementation. Every Week 3.00 3
Directed Reading Non Contact Reading of lecturer-recommended information sources. Every Week 3.00 3
Independent Study Non Contact Independent practical work. Every Week 6.00 6
Total Weekly Learner Workload 12.00
Total Weekly Contact Hours 3.00
 
Module Resources
Recommended Book Resources
  • Lance Eliot. (2018), Lance Eliot 2018, Introduction to Driverless Self-Driving Cars: The Best AI Insider.
  • Packt. (2018), Deep Learning for Computer Vision: Expert techniques to train advanced neural networks using TensorFlow and Keras.
  • Lance Eliot. (2017), New Advances in AI Autonomous Driverless Self-Driving Cars: Artificial Intelligence.
  • Packt. (2017), Machine Learning for OpenCV.
This module does not have any article/paper resources
This module does not have any other resources