EBU7240 Computer Vision
Changjae Oh
Introduction
Semester 1, 2021
Copyright By PowCoder代写 加微信 powcoder
What is coming?
What is missing?
Machines are blind Machine vs Human
Computer Vision in Four Words?
Making computers understand images How simple is that?
Mentimeter
Computer Vision in Four Words?
:: Making computers understand images
• How many people are here?
• Who is a person in the portrait?
• What is this building?
• How is the weather?
• Where is this city?
• What is written?
• Is there any gate?
Computer Vision in Four Words?
:: Making computers understa: nd images
• How many people are here?
• Who is a person in the portrait?
• What is this building?
• How is the weather?
• Where is this city?
• What is written?
• Is there any gate?
→ Object detection
→ Face detection/recognition → Object recognition
→ Scene recognition
→ Place recognition
→ Character recognition
→ Object detection
What is vision?
• What does it mean, to see? “to know what is where by looking”.
• How to discover from images what is present in the world, where things are, what actio
ns are taking place.
• “Vision can be understood as an information processing task which converts a numerical image representation into a symbolic shape-oriented representation.”
Credit: Marr, 1982
Why is vision so hard? • 3D: Viewpoint
Why is vision so hard? • 3D: Viewpoint
Why is vision so hard? • Light
Credit: A. Torralba
Why is vision so hard? • Shadow
Credit: A. Torralba
Why is vision so hard? • Shadow
Why is vision so hard? • Transparent objects
Credit: Hengyi Wang (JP Final project 2021) http://corsmal.eecs.qmul.ac.uk/
Why is vision so hard? • What is this object?
Why is vision so hard? • What is this object?
Why is vision so hard? • What is this object?
It is not just about pixels!
Computer Vision and Image Processing? • Computer Vision begins with Image Processing!
Audio Processing
Signal Processing
Machine Learning Artificial Intelligence Computer Vision
Image Processing
Mathematics
• Computer Vision: techniques for enabling a computer to see a real-world using images as the human being does.
• Image Processing: fundamental techniques for image acquisition, processing, analysis
Computer Vision and
EBU6018 Advanced Transform Methods
EBU7240 Computer
EBU5303 Multimedia Fundamentals
EBU6230 Image and Video Processing
Computer Vision: Low
level Vision
Low-level vision
• Enhancement
• Restoration
• Filtering
• Feature extraction
Mid-level vision
• Grouping
• 3D geometry
• 3D reconstruction • Tracking
High-level vision
• Classification • Detection
• Tracking
• Action & Pose
Computer Vision: Mid
level Vision
Mid-level vision
• Grouping
• 3D geometry
• 3D reconstruction • Tracking
Low-level vision
• Enhancement
• Restoration
• Filtering
• Feature extraction
High-level vision
• Classification • Detection
• Tracking
• Action & Pose
Computer Vision: High
level Vision
Low-level vision
• Enhancement
• Restoration
• Filtering
• Feature extraction
Mid-level vision
• Grouping
• 3D geometry
• 3D reconstruction • Tracking
High-level vision
• Classification • Detection
• Tracking
• Action & Pose
Computer Vision: Deep learning approach • Multiple levels of (learned) representation
EBU7240 Computer Vision
Changjae Oh
Applications
Semester 1, 2020
Why computer vision matters?
Safety Health Security
Comfort Fun Access
Ridiculously brief history of computer vision
• 1966: Minsky assigns computer vision as an undergrad summer project
• 1960’s: interpretation of synthetic worlds
• 1970’s: some progress on interpreting selected images
• 1980’s: ANNs come and go; shift toward geometry and increased mathematical rigor
• 1990’s: face recognition; statistical analysis in vogue
• 2000’s: broader recognition; large annotated datasets
available; video processing starts
• 2010’s: Deep learning with ConvNets
• 2020’s: Widespread autonomous vehicles?
• 2030’s: robot uprising?
Guzman ‘68
Turk and Pentland ‘91
Slide credit: J. Hays
Applications
Motion capture
Microsoft XBox Kinect
Applications
Face recognition
https://www.cl.cam.ac.uk/~jgd1000/afghan.html
Applications
Face Alignment
https://www.maccosmetics.co.uk/virtual-try-on http://jbhuang0604.blogspot.com/2013/04/miss-korea-2013-contestants-face.html
Applications
Smile detection
Slide credit:
Applications
Optical character recognition (OCR)
Slide credit:
Applications
Defect detection
https://deepsystems.ai/solutions/road-defects-detection
Applications
Cleaning robot
Roomba’s new robotic vacuum remembers your home’s layout for quicker cleaning
Applications
Intel Replay Technology Delivers All-Star Experience for MLB Fans
Applications
Object recognition
Applications
3D from mobile phone
Applications
Indoor scene reconstruction
https://youtu.be/Ebyz8xxoK1w
Applications
Video Matting/Composition
Applications
Style Transfer
A Neural Algorithm of Artistic Style [Gatys et al. 2015]
Applications
driving cars
Applications
Healthcare
Why is computer vision timely? • Lots of image data is being collected
Why is computer vision timely?
• Lots of image data is being collected ̶ Bigger data is coming
Why is computer vision timely? • There is a lot of information in images
Why is computer vision timely? • Computer vision is starting to work …
http://www.sony-asia.com/article/271940/section/product/product/dsc-wx1
50 years ago
Slide credit: A. Torralba
10 years ago…
Slide credit: A. Torralba
Next topic
• Let’s talk about a computer’s eye (a.k.a. camera) ̶ Prerequisite
• Review EBU6230 Image/Video Processing – Week2: Image Transformations
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com