EBU7240 Computer Vision
Course Overview
Semester 1, 2021
Changjae Oh
Copyright By PowCoder代写 加微信 powcoder
Course Overview
Unit 1: Early vision / Low-level vision
• Introduction / Camera / Restoration / Feature detection Unit 2: Mid-level vision
• Fitting / Grouping / Calibration / Epipolar /Stereo matching Unit 3: Mid-/High-level vision
• Tracking / Recognition / Detection
Unit 4: Deep learning for computer vision
• Introduction / Loss / Backpropagation / CNN / Deep learning with practice
Course Overview
Unit 1: Early vision / Low-level vision
• Introduction / Camera / Restoration / Feature detection
• Lab1: Setting up – image/video representation in Python
• CT1: Early vision / Low-level vision
Unit 2: Mid-level vision
• Fitting / Grouping / Calibration / Epipolar /Stereo matching
• Lab2: Restoration and features
• CT2: Mid-level vision
Unit 3: Mid-/High-level vision
• Tracking / Recognition / Detection
• Lab3: Fitting and grouping
• Lab4: Tracking and detection + In-lab assessment
• CT3: Mid-/High-level vision
Unit 4: Deep learning for computer vision
• Introduction / Loss / Backpropagation / CNN / Deep learning with practice
• Coursework report submission (Deadline: 23:59, 21st December 2021, UK time)
• CT4: Deep learning for computer vision
Course Details
• Blended Teaching – How?
Module Delivery = 50% live lectures + 50% recorded lectures
̶ Lectures
Mon (3-435)
Telecom_M_G1
Thur (3-437)
19:20-20:05
20:10-20:55
19:20-20:05
Thur (3-535)
Telecom_M_G2
Mon (3-435)
20:10-20:55
16:35-17:20
17:25-18:10
Labs/ Class Tests
16:35-17:20
17:25-18:10
Course Details
• Blended Teaching – How?
Recorded Lectures
To deliver the content in detail
Live lectures
Review the past content + Interactive sessions
Module Delivery
Mon (3-435)
Thur (3-437)
Mon (3-435)
Thur (3-535)
Telecom_M_G1
20:10-20:55
Labs/ Class Tests
19:20-20:05
20:10-20:55
19:20-20:05
Telecom_M_G2
16:35-17:20
17:25-18:10
16:35-17:20
17:25-18:10
Course Details • Recorded Lectures
Recorded Lectures
Recorded Lectures
To deliver the content in detail
Students should take the recorded lecture before the next live session
Mon (3-435)
Thur (3-437)
20:10-20:55
Mon (3-435)
Labs/ Class Tests
Thur (3-535)
Telecom_M_G1
19:20-20:05
20:10-20:55
19:20-20:05
Telecom_M_G2
16:35-17:20
17:25-18:10
16:35-17:20
17:25-18:10
Course Details • Live lectures
Live Lectures
Brief review about past recorded lectures
Interactive sessions using Mentimeter Going through exercises together + Q&A
Mon (3-435)
Thur (3-437)
20:10-20:55
Mon (3-435)
Labs/ Class Tests
Thur (3-535)
Telecom_M_G1
19:20-20:05
20:10-20:55
19:20-20:05
Telecom_M_G2
16:35-17:20
17:25-18:10
16:35-17:20
17:25-18:10
• 4 times: 8th, 10th, 12th, 14th BUPT week
Tuesday afternoon, (13:00 – 14:35)
科研楼 on main (Xitucheng) campus
Telecom_M_Y4_G1 (Room 116),Telecom_M_Y4_G2 (Room 120)
Thur (3-437)
Thur (3-535)
Telecom_M_G1
Mon (3-435)
19:20-20:05
20:10-20:55
19:20-20:05
20:10-20:55
Telecom_M_G2
Mon (3-435)
16:35-17:20
17:25-18:10
Labs/ Class Tests
16:35-17:20
17:25-18:10
Assessment
• Exam (80%)
̶ One written exam
• Coursework (20%)
̶ Individual coursework (15%)
• Development of computer vision tasks • PythonandOpenCV
In-class tests (5%)
Test to be done in each office hour (Four in-class tests)
Each test covers each unit’s content
EasyquestionsusingQMPlus
Top 2 marks (out of 4 tests) will be counted. (2.5% each)
Absence will be marked as zero (NO excuse of your absence will be accepted)
Assessment
• In-lab assessment (30% of individual coursework) ̶ During the Lab4 hours
̶ Assessment of your coursework covered in Lab1-3
• Report (70% of individual coursework)
̶ use the provided layout, with provided guideline
̶ at the end of the semester (Deadline: 21st December 2021)
Individual Coursework (1/3)
One .py code to each problem
Zero mark will be given if the result is not reproducible Zero mark will be given if any unauthorized library is used
Assessment
• In-lab assessment (30%)
̶ will be evaluated by
1) running the implemented codes,
2) checking during the in-lab assessment: the understanding of the tasks with a short conversation with a TA
• Report (70%)
̶ will be evaluated based on
1) the quality of the analysis
2) the discussion of the results obtained in the coursework tasks
Individual Coursework (2/3)
Assessment
̶ A dataset provided from this module (image + video)
→ Quantitative assessment
̶ A dataset collected by yourself (image + video) → Qualitative assessment
Individual Coursework (3/3)
Assessment
• Submit 1) your report and 2) zip file to the QMplus.
Coursework Submission@
QMplus submission example:
• EBU7240_CHANGJAE_OH_1711XXXXX.pdf • EBU7240_CHANGJAE_OH_1711XXXXX.zip
Name the zip file you submit as:
Max size of the zip file: 50M
The outputs of your implementations should be generated in the \results directory
No need to submit the outputs of your code (we will reproduce them!), just make the \results directory
The zip file will contain the following folders:
Deadline: 23:59, 21th Dec 2021 UK time
EBU7240_FIRSTNAME_FAMILYNAME_QMSTUDNETNUMBER ├── inputs
├── results
• Four in-class tests
class Test
To be done in each office hour
Less than 20 min
Students should be in the classroom (Scores will be accepted ONLY WHEN attendance is recorded)
Easy online-test using QMPlus Each test covers each unit’s content
QuestionsforTelecom_M_G1andTelecom_M_G2willbedifferent
For a fairness issues, ONLY the problems used in the live-session will be covered (with modification)
Top 2 marks will be counted. (2.5% each) Absence will be marked as zero
Any excuse of your absence will NOT be accepted
Office Hours • When?
̶ During office hours (OH), but after the class test (<20 min) • Where?
I will post the meeting link through QMPlus
Anyone can drop in with his/her own MS Teams account and have a video meeting
A few tips
• Define, define, define!
̶ ex) EBU6230- Image Video processing
Opening: M·S=(M S)⊕S
A few tips
• There are several traps to prevent your plagiarism ̶ Don’t copy others
̶ You’ll need to create your own dataset ̶ ex)
Coursework
By the end of this module, you will
• understand fundamental tasks involved in computer vision tasks
• understand the principle of deep learning in computer vision
• become familiar with
̶ the various important techniques in computer vision tasks ̶ Python and OpenCV
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com