Page 353 - AI Ver 3.0 class 10_Flipbook
P. 353
First, we have the input layer which receives the input in several different formats provided by the programmer and
feeds it to the neural network. Minimal processing occurs in the input layer, as it simply passes the raw input data
forward. The output layer produces the final prediction or decision based on the learned patterns. The output at
each node is called its activation or node value.
What is Convolutional Neural Network (CNN)?
Convolutional Neural Network is a type of Artificial Neural Network and is made up of neurons that help in image
recognition and image processing. It uses a deep learning algorithm that takes an input image, processes it by
assigning learnable weights and biases to various aspects/objects in the image, enabling the network to identify
patterns and features helping the system differentiate one image from the other with maximum accuracy. CNNs
reduce the spatial dimensions (size) of the input through operations like pooling, while retaining the essential
features and give the predicted class probabilities for the input image. They are trained to identify and extract the
best features from the images.
CARGO SHIP
STEAM BOAT
CRUISE Cargo Ship 17%
Steam Boat 8%
FISHING BOAT Cruise 75%
INPUT CONVOLUTION - ReLU POOLING CONVOLUTION - ReLU POOLING FLATTEN FULLY SOFTMAX
CONNECTED
Output Probability
FEATURE LEARNING CLASSIFICATION
Input Image
Image Processed by CNN
In the above diagram, an input image is provided, processed through a CNN, and a prediction is generated based
on the labels in the corresponding dataset.
Brainy Fact
Convolutional neural networks (ConvNets) were first introduced in the 1980s by Yann LeCun, a computer
science researcher. Its early version called LeNet (after LeCun), were used to recognise handwritten digits. It
found its use in postal services to read zip codes on envelopes and in banking/financial sectors to read digits
on cheques.
Layers of Convolutional Neural Network (CNN)
The different layers of a Convolutional Neural Network (CNN) are shown in the following figure:
CARGO SHIP
STEAM BOAT
CRUISE
FISHING BOAT
INPUT CONVOLUTION - ReLU POOLING CONVOLUTION - ReLU POOLING FLATTEN FULLY SOFTMAX
CONNECTED
FEATURE LEARNING CLASSIFICATION
Computer Vision (Practical) 351

