Registry
Module Specifications
Archived Version 2020 - 2021
| |||||||||||||||||||||||||||||||||||||||||
Description Computer vision applications have significantly expanded over the last decade and this core skill set is always in high demand by employers. This module will build on the basic concepts with a view to delving deeper into core computer vision, machine learning and deep learning topics. As well as examining traditional computer vision concepts (i.e. feature extraction and machine learning) a key focus of the module will be on deep learning as applied to computer vision. We will examine the core concepts behind deep learning for computer vision with a specific focus on Convolutional Neural Networks (CNN). Students will learn how to design and tune such networks in a range of practical applications and assignments. In addition we will examine a range of deep learning architectures ranging from AlexNet upto the current state of the art in this ever expanding field. Deep learning based computer vision forms the core of many of the recent developments in this field and has been widely adopted as a core AI tool by all the key industrial players such as Google, Facebook, IBM, Apple, Baidu ... as well as a wide range of highly innovative startups. All computer vision and deep learning concepts will be reinforced by guided practical work and case studies. This module is primarily aimed at those who aim to undertake research in computer vision or require a deeper understanding of the subject to address commercial computer vision development. Computer vision applications span a wide range of disciplines including industrial/machine vision, video data processing, biomedical engineering, healthcare, astronomy, imaging science, sensor technology, multimedia and enhanced reality systems. Please refer to the modules summary syllabus for a breakdown of the course content. This module will require basic programming skills. See the EE544 module website (http://www.eeng.dcu.ie/~whelanp/ee544/) for details on the computer vision & deep learning development environment. | |||||||||||||||||||||||||||||||||||||||||
Learning Outcomes 1. Recall, review and analyse the advanced theories, algorithms, methodologies and techniques involved in traditional and deep learning based computer vision. 2. Illustrate their ability to comprehend and interpret issues relating to the design of advanced traditional and deep learning based computer vision. 3. Synthesize and evaluate the relevant merits of competing advanced computer vision techniques. 4. Apply computer vision techniques in a range of application scenarios. 5. Develop an deep understanding of the issues involved in the evaluating computer vision system implementation. 6. Demonstrate the ability to implement a computer vision pipeline. | |||||||||||||||||||||||||||||||||||||||||
All module information is indicative and subject to change. For further information,students are advised to refer to the University's Marks and Standards and Programme Specific Regulations at: http://www.dcu.ie/registry/examinations/index.shtml |
|||||||||||||||||||||||||||||||||||||||||
Indicative Content and
Learning Activities Introduction• Introduction to Python Computer Vision Development Environment • Computer vision Pipeline • Traditional vs deep learning approaches to computer vision • Learning Outcomes • Module Protocol • Assessment Requirement • Support Material & Website • Software Tools • Case StudiesInterest Point Detection & Feature Extraction• SIFT - Scale Invariant Feature Transform • Histogram of Oriented Gradients (HOG) • Deformable Part Models (DPM)Machine Learning for Computer Vision• Classification • Feature Normalisation • Evaluation of Classifier Performance • Non-Parametric Classifiers / Decision Trees (DT) • Support Vector Machine (SVM) • SVM Multi-class ClassificationDeep Learning for Computer Vision (ANN & CNN)• Artificial Neural Networks • Logistic (Linear) Classifier • Gradient Descent / Stochastic Gradient Descent (SGD) • Backward Propagation • Regularisation Methods • Supervised Deep Learning • Convolution Neural Networks • Transfer Learning • Architectures • Unsupervised LearningDeep Learning for Computer Vision (Classification, Visualisation & Localisation)• CNN classification • Data Augmentation • Visualising CNN filters • Localise Objects with Regression • Object Detection as Classification • Region-Based CNN (R-CNN) • Single Shot Detectors (SSD)Deep Learning for Computer Vision (Segmentation, Detection & Advanced)• Semantic Segmentation • Fully Convolutional Networks • Instance Segmentation • Style Transfer Network • Deep Dream • Generative Adversarial NetworksMotion• Optical Flow: 2D • Optical Flow Constraints • Local / Global Approaches • Feature Matching – Motion Correspondence • Kanade Lucas Tomasi (KLT) Tracking | |||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||
Indicative Reading List
| |||||||||||||||||||||||||||||||||||||||||
Other Resources 30461, Module Website, 0, EE544, http://www.eeng.dcu.ie/~whelanp/ee544/ee544_notes.html, 30462, Lecturer Website, 0, paulwhelan.eu, http://paulwhelan.eu/, | |||||||||||||||||||||||||||||||||||||||||
Programme or List of Programmes | |||||||||||||||||||||||||||||||||||||||||
Archives: |
|