Advanced Vision Module

Lecture Notes ]   [ Announcements ]   [ Assignments ]   [ MATLAB ]   [ Books ]   [ FAQ ]  

Introduction

The goal of this course is provide you with the skills to understand and sketch out solutions to a variety of computer vision applications. You should end up with the skills to tackle novel situations and incompletely defined applications. We will approach this by looking at 6 simplified computer vision systems that cover a large portion of the range of both applied and research computer vision. 10 Minute Introduction

This module assumes the students have a secondary school understanding of geometry, matrix algebra, trigonometry, physics and programming concepts. Knowledge of elementary optics, signals and photography would be helpful. Students must be able to program and be able to work in small teams. Normally, a student will have attended Introduction to Vision and Robotics (IVR) before attending this module, but exceptions can be made if you have the necessary background. All of the assignments will be in Matlab. This was introduced in IVR, and some extensions will be introduced here. Please ensure familiarity with the concepts, techniques and practical aspects of the vision component of the IVR course. See the AV Learn page for links to Matlab tutorials.

Lecturer

Bob Fisher, Bayes 1.11, rbf @ inf.ed.ac.uk, (0131) 651-3441
Before emailing with a question - please check the module FAQ.

Demonstrators

Hanz Cuevas Velasquez (s1678460)

Class Meetings:

Semester 2 full class meeting times and rooms: Monday and Thursday 2:10 pm (Venues: Monday: Lister Institute G.01, Thursday DHT Lower Ground room 9).

Lab demonstration sessions: Starting week 2, Monday 4-5, Tuesday 4-5, Thursday 4-5, or Friday 4-5 (just come to 1 hour) in Appleton Tower room 5.08 South Lab.
There are 5 non-assessed lab exercises that will help you develop your matlab and image handling skills. See the 5 drill exercises in Learn. See also the AV Learn page for links to Matlab tutorials.

Lecture Plan

IMPORTANT INFORMATION: This course is not taught by the traditional lectures. Instead, AV uses an Inverted Classroom method. This means that you will have about 15 hours of video to watch in your own time. This material is assessible. There are still 2 full class meetings each week. Each class will consist of 2 parts: 1) discussing any questions about the videos that you either suggest in advance or raise in class on the day. 2) There will be a simple non-assessed groupwork exercise to explore the issues raised in the videos that you have just watched.

Here is the link to the lecture videos, associated readings and associated Matlab in the University's LEARN system. You will need an EASE account to access this materials and then use the 'Login' button at the upper right of the screen. Then click on the large Login with EASE button. After that click on "Advanced Vision (Level 11) (2018-2019)[SV1-SEM2]". You should: (a) Read the introduction "Summary of teaching materials and approach". (b) Watch the materials for the corresponding week (as given in the table below) by the Monday class. (c) Get access to the materials by clicking, for example, on 'Course introduction and review' (left edge) -> '2. Coordinate geometry transformation review' (right panel). This exposes the materials. (d) Read the lesson plan, download the PDF slides to be annotated while you watch the video, and then try to answer the Review Question.

Here is an Introduction to how the course is intended to be followed.

Here is a proposed schedule of video watching and guest lectures:

WeekVideosMondayThursday
Jan 14 Introduction: modules 1-3; Visual Ethics; Flat parts System 1: modules 1-3 Course Intro Q&A + drill
Jan 21 Flat parts System 1: modules 4-6 Detection and tracking System 3: modules 1-4 Q&A + drill Q&A + drill
Jan 28 Detection and tracking System 3: modules 5-12 Q&A + drill Q&A + drill
Feb 4 Range Image Analysis System 4: modules 1-8 Q&A + drill Q&A + drill
Feb 11 Stereo based 3D part recognition System 6: modules 1-5 Q&A + drill Q&A + drill
Feb 18 Festival of Creative Learning Week no class no class
Feb 25 Stereo based 3D part recognition System 6: modules 6-11 Q&A + drill Q&A + drill
Mar 4 Deforming flat part recognition System 2: modules 1-8 Q&A + drill Q&A + drill
Mar 11 Persistent tracking and behavior recognition System 5: modules 1-6 Q&A + drill Q&A + drill
Mar 18 Deep Nets for Vision Modules 1-4 + Course Summary Q&A + drill Previous Exam Review

News

  Latest information, class announcements, handouts, etc

Activities

  1. About 15 hours of lectures on the above syllabus are given, depending on the speed of delivery.
  2. Some supplementary readings that explain the lecture content in a different way.
  3. 1 assessed laboratory practical exercise: 30 hours for 30% of course mark. These laboratory exercises are done in teams of 2 to encourage development of team skills: teamwork exercises skills desired by employers and improves the learning process by encouraging discussion of topics.
  4. Outside reading and exam revision: 15 hours.
  5. Total: 100 hours

Assessment

A 2 hour examination in the late spring accounts for 70% of the module mark and the practical work accounts for the other 30%. You can see past exam papers. Go to: https://exampapers.ed.ac.uk/, enter "Advanced Vision" and select "Advanced Vision (Level 11)". Click "Search".

See the Learn pages for the coursework information.

Discussion Forum

The AV course has a Piazza Discussion Forum. Please use this for question asking and discussion as you will probably get a faster answer from your colleagues and the TA than by direct email to Bob Fisher. Most students have been enrolled for the AV Piazza Forum. If you have not been enrolled, email rbf@inf.ed.ac.uk.

Other Information


This page is maintained by the course lecturer, Bob Fisher, rbf@inf.ed.ac.uk, room IF 1.26, ext 513441.


Home : Teaching : Courses 

Informatics Forum, 10 Crichton Street, Edinburgh, EH8 9AB, Scotland, UK
Tel: +44 131 651 5661, Fax: +44 131 651 1426, E-mail: school-office@inf.ed.ac.uk
Please contact our webadmin with any comments or corrections. Logging and Cookies
Unless explicitly stated otherwise, all material is copyright © The University of Edinburgh