Computer Vision - CNN for image classification

One of the most important uses of self-driving automobiles is image categorization.

Where the semi and completely autonomous car industries are rapidly expanding, we now see self-driving cars from the likes of Tesla and Kia moving around everywhere.

We will be exploring the technology behind this and how computers see and classify images, from understanding waste types to identifying humans on the road against inanimate objects and signs.

We will build a traffic sign classification model using CNN.

Transcript

1. Computer Vision -
   CNN for Image
   Classification
   by Wale Ayandiran
   Presented at AI Wales meetup, Tramshed Tech, Cardiff.
   08-06-2022
   

   View Slide

2. We might start by identifying certain pictures of animals
   ourselves. Shall we?
   How does computer identify images
   with their labels?
   

   View Slide

3. Can you tell these wild cats aparts?
   

   View Slide

4. Leopard Cheetah
   Jaguar Tiger
   Image credit: Google
   

   View Slide

5. Machine
   Learning
   

   View Slide

6. Lion Cat
   Wolf
   Dog
   Fox
   

   View Slide

7. Cat Lion
   Dog
   Arctic Fox
   Wolf
   Image credit: Google
   

   View Slide

8. How are we as
   humans able to
   differentiate things
   easily?
   

   View Slide

9. Recognizing Patterns
   How are we as humans able to
   differentiate things easily?
   

   View Slide

10. Credit: MIT
    Feature Detection
    Low level feature detection to high level feature detection
    

    View Slide

11. What do computers see?
    

    View Slide

12. Images are matrix of
    numbers
    

    View Slide

13. Images are matrix of numbers What computers see
    

    View Slide

14. Images are matrix of numbers
    What computers see
    Credit: https://openframeworks.cc/ofBook/chapters/image_processing_computer_vision.html
    

    View Slide

15. Credit: https://www.keithbcarter.com/think-differently/how-a-machine-sees
    Computer detects features of an image as matrix of
    numbers.
    

    View Slide

16. How do we make computers
    identify images?
    Neural Networks.
    Neural networks are computing
    systems with interconnected nodes
    that work much like neurons in the
    human brain. Using algorithms, they
    can recognize hidden patterns and
    correlations in raw data, cluster and
    classify it, and – over time –
    continuously learn and improve.
    Credit: https://www.sas.com/en_us/insights/analytics/neural-networks.html
    

    View Slide

17. A simple neural network includes an
    input layer, an output (or target) layer
    and, in between, a hidden layer. The
    layers are connected via nodes, and
    these connections form a “network” –
    the neural network – of interconnected
    nodes.
    How Neural Network Works
    Credit: https://www.sas.com/en_us/insights/analytics/neural-networks.html
    

    View Slide

18. ● Artificial Neural Network (ANN)
    ● Convolutional Neural Network (CNN)
    ● Recurrent Neural Network (RNN)
    3 Popular Neural Networks
    

    View Slide

19. Convolutional
    Neural Networks
    

    View Slide

20. Learning Visual Features
    ➔ Filter extraction
    Applying filters and generating
    smaller pieces of image from the
    original image.
    ➔ Features Mapping
    Using the filters extracted to
    create a feature map of the
    image.
    

    View Slide

21. Image credit: MIT
    

    View Slide

22. Tiger
    Tiger’s eyes ? = Y
    Tiger’s nose ? = Y
    Tiger’s mouth/whiskers ? = Y
    Tiger’s paws ? = Y
    Tiger’s tail ? = Y
    Tiger’s head ? = Y
    Tiger’s body ? = Y
    Is it a Tiger ? = Y
    Image credit: Google
    

    View Slide

23. Main parts of a CNN
    Three main parts to a CNN
    ➔ Convolution
    Application of filters to generate feature
    maps
    ➔ Non-linearity (ReLU)
    Replacing negative values with zero (0)
    often ReLU (Rectified Linear Unit)
    ➔ Pooling
    Reduces dimensions and distortions,
    making the model tolerant towards
    image variations.
    

    View Slide

24. Feature Extraction Classification
    Image credit: codebasics
    

    View Slide

25. Some limitations of CNN
    in Image classification
    ❏ Requires large datasets to train
    ❏ Can take a lot of time to train if there are several layers
    ❏ Can be computing power intensive.
    

    View Slide

26. A CNN model to identify traffic signs
    https:/
    /bit.ly/traffic_cnn
    

    View Slide

27. Thank You!
    Find slide and code link below
    Demo link:
    https:/
    /bit.ly/traffic_cnn
    Github
    https:/
    /github.com/waleCloud/solid-fortnight
    

    View Slide

28. Wale Ayandiran.
    Software and AI Engineer.
    Find me at https://walecloud.me
    Twitter @walecloud
    

    View Slide