Hide menu

TSBB17 Visual Object Recognition and Detection HT2020

This course is worth 6 ECTS credits, which corresponds to approximately 160h of work per student. The time is divided among the following activities:

The course is given during the HT1 period, and will run in "distance mode". Lectures will be made available as pre-recorded videos that can be accessed through the Lisam course page. Scheduled lecture slots will be used as Q&A sessions over Zoom. Note that not all the time slots in the TimeEdit schedule will be used. A few spare slots have been reserved in case a session has to be moved. The actually planned lectures are listed below on this page, and any changes to this plan will be announced over email, during the lectures, and on this page.

People

Our offices are in the B-building. Lecturers sit in corridor D, first floor between entrances 25 and 27. Project supervisors sit in Visionen.

Documents

  • Detailed information and documents related to the two projects will be provided through the course repository hosted in LiUs GitLab (accessible to anyone with a LiU login).
  • Course description in Studieinfo (the web interface to Bilda)

Literature

  • Goodfellow et al., Deep Learning, MIT Press (2016).
    The book is available as full pdf. See also Book webpage.
  • Richard Szeliski, Computer Vision: Algorithms and Applications, Springer Verlag (2011).
    The book is available as an on campus e-book via the LiU library. See also Book webpage.
Additional material for side reading will be made available in the course repository.

Registration

If you intend to take the course but are not registered, make sure to register ASAP, using the Student portal. You need to be registered on the course to receive course email, and to have results input to Ladok. If you take the course but are not registered to any program at the University, please contact the course examiner in order to make sure that you receive email about the course.

Examination

  • The course has a written examination in English. This year, the exam will be a "distance exam", which is accessed, and handed in using the Submission tab on the Lisam page for the course. The exam has the same format as previous years. See examples below:
    • Example exam from 2017-10-17 [pdf]
    • Example solutions from 2018-01-02 [pdf]
    • The next examination opportunity is at the end of the course, 2020-10-24.
    • Note that last year's 2nd reexamination, 2020-08-28 is cancelled, as noone registered for it.
  • The projects are examined using written reports and oral presentations in English.

Lecture schedule 2020

During HT1 2020 the lectures are given in the form of pre-recorded videos, with a Q&A session at the scheduled lecture time (a dedicated time when you can ask questions about the content). For best use of your time, you should thus watch the lecture before the time in the schedule, and note down your questions. The pre-recorded lectures can be accessed through the Lisam page for the course.

Before the lectures, the lecture slides from last year can be found in the course repository. Updated slides will be pushed after the video content has been produced. The repository also contains additional relevant literature for side reading.

Date,Time Activity Teacher
August 31: 08.15-10
Q&A starts 9.15
Lecture 1 Q&A
Introduction
Per-Erik Forssén
September 1: 10.15-11
Lecture 2 Q&A
Feature Descriptors and Evaluation
Per-Erik Forssén
September 2: 13.15-14
Lecture 3 Q&A
Convolutional Neural Networks: Introduction and Theory
Michael Felsberg
September 4: 9.00-9.45
Q&A starts 9.00
Lecture 4 Q&A
Image Classification with Convolutional Neural Networks
Michael Felsberg
September 8: 10.15-12
Lecture 5 Q&A
Compound Descriptors and Metric Learning
Per-Erik Forssén
September 9: 13.15-15
Lecture 6
Project 1: Visual Object Recognition
Per-Erik Forssén
September 15: 10-12
Q&A starts 10.15
Lecture 7 Q&A
Visual Object Detection
Per-Erik Forssén
September 16: 13-15
Q&A starts 14.00
Lecture 8 Q&A
Visual Object Tracking: Introduction
Michael Felsberg
September 22: 10.15-12
Q&A starts 10.45
Lecture 9 Q&A
Discriminative Correlation Filters for Visual Tracking
Michael Felsberg
September 23: 13.15-15
Seminar 1
Presentation of project 1
Per-Erik Forssén
September 23: 15.15-17
Lecture 10
Project 2: Visual Object Tracking
Per-Erik Forssén
October 14: 13.15-15
Seminar 2
Presentation of project 2
Per-Erik Forssén
A more extensive schedule can be found in TimeEdit. It also contains scheduled project time ("Projekttid", when you have exclusive access to the computer room Olympen) as well as backup lectures. Note that teachers will only be present at activities listed in the table above.

Projects

The projects are conducted in groups of 4 or 3 students (in order of preference).

  • Project 1: Visual Object Recognition
    Introductory lecture: September 9: 13.15-15, Zoom
    Report due: September 20 (a Sunday)
    Presentation seminar: September 23: 13.15-15, Zoom
  • Project 2: Visual Object Tracking
    Introductory lecture: September 23: 15.15-17, Zoom
    Report due: October 7
    Presentation seminar: October 14: 13.15-15, Zoom

General resources

We recommend using the following software:

  • PyTorch A deep learning framework for Python (support code for the projects is written for PyTorch).
  • PyCharm A Python IDE (also installed in Olympen).
Other related software of interest:
  • MatConvNet Deep Learning framework for Matlab.
  • Caffe Deep Learning library.
  • Theano Deep Learning library.
  • VLFeat has a a useful code library, both for Matlab and C/C++. For example, there is an alternative implementation of SIFT here, and also an implementation of MSER here. Both are made by Andrea Vedaldi.
  • The Visual Geometry Group at Oxford University maintains code for affine invariant region detectors, produced in cooperation with other groups.
  • LIBSVM A Library for Support Vector Machines (Matlab, Python).

Project repositories

Project code should be developed under versioning control, with changes tracked according to LiU-ID of the participating group members.

  • Project groups should get their repositories from GITLab at LiU Note: this is not GitHub, and GitHub should not be used.
  • Alternatively, the IDA GITLab may also be used.


Last updated: 2020-09-21