Week 5 - Higher-Level Computer Vision

Question 1

What are the 4 main Recognition problems?

Answer 1

• Recognition: Identifying the main object in an image
• Detection: Find the location of all objects
• Segmentation: Assign all pixels to objects
• Pose: Find the location of the object parts

Question 2

What is the formal definition of the ‘Detection’ problem in Recognition problems?

Answer 2

Find the location of all objects in the scenes in terms of providing a bounding box

Question 3

What is Semantic Image Segmentation?

Answer 3

It’s the process of partitioning the image into ‘meaningful’ segments
You group pixels based on ‘common’ properties

Question 4

What is Instance Image Segmentation used for?

Answer 4

If you need to differentiate different instances of the same object

Question 5

What three main features are required for a good Object Recognition Model?

Answer 5

• Data: Images containing objects from that class and images from all other classes
• Feature Extraction: Work with features extracted from images
• Machine Learning: From the features extracted, initiate and train a model that recognises this particular object class

Question 6

How do HOG features work?

Answer 6

• Divide image into a grid of cells e.g. 8x8
• Compute edges and their orientation for every pixel location
• Compute histogram of gradient orientations in each cell

Question 7

What is Bag of Features?

Answer 7

Bag of Features methods analyse the large set of very specific features generated by a training set of images and identify a small set of useful, more generic features

Question 8

How does Object Recognition work with Bag of Features?

Answer 8

• Take a bunch of images: Extract features, build up a ‘dictionary’ of common features
• Then, given a new image, extract features:
  – For each feature, find the closest visual word in the dictionary
  – Build a histogram to represent the image

Question 9

How does Viola-Jones Recognition work in practice?

Answer 9

• Slide a window across the image and evaluate a face model at every location

Question 10

What are the key ideas that were pulled from Viola-Jones Recognition?

Answer 10

• Integral images for fast feature evaluation
• Boosting for feature selection
• Attentional cascade for fast rejection of non-face windows

Question 11

How does the Integral Image work?

Answer 11

The integral image computes a value at each pixel that is the sum of the pixel values above and to the left of the source pixel inclusive.

Question 12

How is Feature Extraction performed using Integral Images?

Answer 12

Features are extracted from sub-windows of a sample window, and each of the four feature types are scaled and shifted across all possible combinations.

Question 13

What is the formal definition of Boosting?

Answer 13

Boosting is a classification scheme that works by combining weak learners into a more accurate ensemble classifier, where a weak learner is defined as a learner that does only slightly better than random chance.

Question 14

How does Boosting work?

Answer 14

• Need a training set of labelled examples
• Start with all examples equally weighted
• Learn a series of recognition rules
• Re-weight examples so incorrect recognition by nth classifier makes that example more important to the n+1th
• No single rule/classifier can separate complex objects from complex backgrounds, but a combination can
• Weights are determined automatically

Question 15

What is the Classical Approach in Computer Vision?

Answer 15

Apply learned operations to user-defined features:
- Design/choose features
- Design/choose a classifier
- Train the classifier

Question 16

How does Bag of Words reduce reliance on the user?

Answer 16

It clusters the results of applying the user-defined set of feature-detection operators to form a more generic visual vocabulary.