Computer Vision

Cod y Modiwl: CS34020
Teitl y Modiwl: Computer Vision
Blwyddyn Academaidd: 2027/2028
Semester: Semester 1
Cyd-gysylltydd y Modiwl: Professor Bernard Tiddeman
Rhestr Ddarllen: Gweld ar Aspire
Rhagofynion: CC12320 neu CS12320
Anghymharus (Unrhyw Flwyddyn Acad): Replacement for CS34110
Staff Eraill sy'n Cyfrannu: Dr Wayne Aubrey

Dulliau Asesu

Math o Asesiad	Manylion Asesiad	Cyfran
Asesiad Semester	Computer vision assignment: 30 Awr	50%
Arholiad Semester	Written Exam: 2 Awr Written exam, with limited open notes, based on a topic indicated in advance.	50%
Asesiad Ailsefyll	Computer vision assignment: 30 Awr	50%
Arholiad Ailsefyll	Supplementary Exam: 2 Awr Will take the same form, under the terms of the Department's policy.	50%

Canlyniadau Dysgu

Wedi cwblhau'r modiwl dylai'r myfyrwyr fedru:

Express a consolidated and extended understanding and knowledge of Computer Vision techniques.
Compare, critically evaluate and discuss competing methods.
Explain the problems, techniques and difficulties associated with the different areas of Computer Vision.

Disgrifiad cryno

The module will introduce the subject of Computer Vision with applications in a variety of contexts, including robotics, security and image analysis. It will start with low-level vision such as edge detection, feature detection, and segmentation. Intermediate vision will describe various techniques to infer 3 dimensional information from images. Some high-level techniques will be introduced, leading to discussion of deep learning approaches.

Nod

The aim of the module is provide a grounding in applied computer vision, including real-time and low-level techniques through to high-level and deep learning approaches. It is intending to prepare students for major projects and employment opportunities utilising computer vision techniques.

Cynnwys

Foundations of vision: Image acquisition, sources of noise, human visual perception, and the evaluation and design of visual computing systems.

Edges and features: The image as landscape, edge detection, feature detection and representation, appearance as feature.

Motion: The video as a 3D dataset, feature tracking, background subtraction, modelling motion and change.

Objects: Grouping features, grouping motion, modelling variability. Learning models.

3D: Shape from X (shading, defocus, occlusion, photometric stereo). Multiview techniques (binocular, structure from motion). Direct 3D capture techniques (lidar, sonar).

Deep learning: Convolutional neural networks, attention mechanisms and Transformer architectures, generative networks, loss functions and training schemes.

Sgiliau Modiwl

Math o Sgiliau	Manylion Sgiliau
Cyfathrebu	Exam writing develops written communication skills
Datblygu personol a chynllunio gyrfa	There is a substantial demand for computer vision expertise.
Datrys Problemau	Problem solving is intrinsic to computing in general.
Gwaith Tim	This module will require individual rather than team work.
Gwella dysgu a pherfformiad ei hun	Independent learning is necessary to complete the module
Rhifedd	Computer vision involves higher level mathematical concepts
Sgiliau pwnc penodol	Computer vision.
Sgiliau ymchwil	This is a research driven module; research skills will be exercised throughout
Technoleg Gwybodaeth	Information and communications technology is intrinsic to computer science.

Nodau

Mae'r modiwl hwn yn cydymffurfio a FfCChC Lefel 6