Scribd Logo
Upload

Welcome to Scribd!

  • 13 views

    8 CNN Example

    Uploaded by

    2K20SE159VipulChauhan
    The document discusses convolutional neural networks (CNNs). CNNs are a type of neural network that can process data with a known, grid-like topology such as image data. CNNs apply successive layers of convolution and pooling operations to extract hierarchical features from input data. The document outlines the key components of CNNs including convolutional layers, activation layers, pooling layers, and fully connected layers. It also provides an example of implementing a CNN using TensorFlow to classify images from the CIFAR-10 dataset.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    8 CNN Example

    Uploaded by

    2K20SE159VipulChauhan
    13 views33 pages
    The document discusses convolutional neural networks (CNNs). CNNs are a type of neural network that can process data with a known, grid-like topology such as image data. CNNs apply successive layers of convolution and pooling operations to extract hierarchical features from input data. The document outlines the key components of CNNs including convolutional layers, activation layers, pooling layers, and fully connected layers. It also provides an example of implementing a CNN using TensorFlow to classify images from the CIFAR-10 dataset.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses convolutional neural networks (CNNs). CNNs are a type of neural network that can process data with a known, grid-like topology such as image data. CNNs apply successive layers of convolution and pooling operations to extract hierarchical features from input data. The document outlines the key components of CNNs including convolutional layers, activation layers, pooling layers, and fully connected layers. It also provides an example of implementing a CNN using TensorFlow to classify images from the CIFAR-10 dataset.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    13 views33 pages

    8 CNN Example

    Uploaded by

    2K20SE159VipulChauhan
    The document discusses convolutional neural networks (CNNs). CNNs are a type of neural network that can process data with a known, grid-like topology such as image data. CNNs apply successive layers of convolution and pooling operations to extract hierarchical features from input data. The document outlines the key components of CNNs including convolutional layers, activation layers, pooling layers, and fully connected layers. It also provides an example of implementing a CNN using TensorFlow to classify images from the CIFAR-10 dataset.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 33

    Convolutional Neural

    Networks_Ex
    Dr. Dinesh Kumar Vishwakarma
    Professor, Department of Information Technology
    Delhi Technological University, Delhi
    Email: [email protected]
    Web page: http://www.dtu.ac.in/Web/Departments/InformationTechnology/faculty/dkvishwakarma.php
    Biometric Research Laboratory
    http://www.dtu.ac.in/Web/Departments/InformationTechnology/lab_and_infra/bml/

    3/7/2023 1
    What is CNN?

    3/7/2023 2
    CNN Benefits
    • Equivariant activities: Replicated features do not make the neural
    activities invariant to translation. The activities are equivariant.

    Representation Translated
    by active representation
    neurons

    translated
    image
    image

    • Invariant Knowledge: If a feature is useful in some locations during


    training, detectors for that feature will be available in all locations during
    testing.
    3/7/2023 Courtesy: Geoffrey Hinton CSC2535 3
    How CNN Works?

    3/7/2023 4
    Possible Case

    3/7/2023 5
    How CNN Works?

    3/7/2023 6
    How CNN Works?

    3/7/2023 7
    How CNN Works?

    Gives an idea

    3/7/2023 8
    How CNN Works?

    First Filter

    Second Filter

    Third Filter

    3/7/2023 9
    How CNN Works?

    3/7/2023 10
    How CNN Works?

    3/7/2023 11
    Convolutional Layer

    3/7/2023 12
    Operations Involved in Convolution Layer
    • Move filter over the image at every possible position.
    • Multiply each image pixel by corresponding feature pixel.

    Similarly

    3/7/2023 13
    Operations Involved in Convolution Layer…
    O/P of
    Conv
    Layer
    for first
    filter
    • Adding all the multiplied values of filter at every move.
    • Divide the sum by total number of pixels.

    3/7/2023 14
    Output of Convolutional Layer

    3/7/2023 15
    Activation Layer: ReLU

    3/7/2023 16
    ReLU: Removes the Negative Values

    3rd Feature 2nd Feature 1st Feature


    First Feature

    3/7/2023 17
    Pooling Layer

    3/7/2023 18
    Max Pooling of size: 2x2, Stride : 2

    4x4
    .77
    7x7
    3/7/2023 19
    Max Pooling of size: 2x2, Stride : 2

    .77

    7x7 4x4

    7x7 4x4

    4x4
    7x7
    3/7/2023 20
    Stacking of Layers
    Convolutional Layer

    i-th layer of CNN

    3/7/2023 21
    Practical CNN Model
    • Most of the practical CNN models have more number of
    layers.

    2x2

    2x2

    2x2

    3/7/2023 22
    Fully Connected Layer
    • Actual Classification Occurs at FC layer

    Flattened the
    final output
    layer

    3/7/2023 23
    Output

    1 1
    2 2
    3 3
    4 4
    5 5
    6 6
    7
    “1” at
    7
    8 8
    9 9
    10 10
    11 11
    12 12

    3/7/2023 24
    Prediction

    1
    Now Network
    2

    3
    is training
    4
    and gives
    5
    prediction for
    6 a Test sample
    7

    10 Is is “X” or “O”?
    11

    12
    Lets Compare
    3/7/2023 25
    Prediction: Compare with ‘X’ and ‘O’
    1 1

    2 2

    3 3

    4 4

    5 5

    6 6

    7 7

    8 8

    9 9

    10 10

    11 11

    12 12

    For ‘X’

    3/7/2023 26
    Prediction: Compare with ‘X’ and ‘O’
    1 1

    2 2

    3 3

    4 4

    5 5

    6 6

    7 7

    8 8

    9 9

    10 10

    11 11

    12 12

    For ‘O’

    3/7/2023 27
    Result: Compare with ‘X’ and ‘O’

    What is result?

    Classified
    as “X”

    3/7/2023 28
    CNN for Edge Detection

    3/7/2023 29
    Implementation of CNN
    The CIFAR-
    10 dataset:
    This dataset
    consists of
    60000 32x32
    colour images
    in 10 classes,
    with 6000
    images per
    class.

    3/7/2023 30
    Model of CNN

    3/7/2023 31
    Implementation in TensorFlow
    def createModel():
    model = Sequential()
    # The first two layers with 32 filters of window size 3x3
    model.add(Conv2D(32, (3, 3), padding='same', activation='relu', input_shape=input_shape))
    model.add(Conv2D(32, (3, 3), activation='relu'))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Dropout(0.25))

    model.add(Conv2D(64, (3, 3), padding='same', activation='relu'))


    model.add(Conv2D(64, (3, 3), activation='relu'))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Dropout(0.25))

    model.add(Conv2D(64, (3, 3), padding='same', activation='relu'))


    model.add(Conv2D(64, (3, 3), activation='relu'))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Dropout(0.25))

    model.add(Flatten())
    model.add(Dense(512, activation='relu'))
    model.add(Dropout(0.5))
    model.add(Dense(nClasses, activation='softmax'))
    return model

    3/7/2023 32
    References
    • https://www.youtube.com/watch?v=umGJ30-15_A

    3/7/2023 33

    You might also like