User at 17:58, 4 September 2023

2023-09-04T17:58:14Z

User: Created page with "A '''Convolutional Neural Network (ConvNet/CNN)''' is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various a..."

2021-12-03T01:52:24Z

Created page with "A '''Convolutional Neural Network (ConvNet/CNN)''' is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various a..."

New page

A '''Convolutional Neural Network (ConvNet/CNN)''' is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various aspects/objects in the image and be able to differentiate one from the other. The pre-processing required in a ConvNet is much lower as compared to other classification algorithms. While in primitive methods filters are hand-engineered, with enough training, ConvNets have the ability to learn these filters/characteristics. The architecture of a ConvNet is analogous to that of the connectivity pattern of Neurons in the Human Brain and was inspired by the organization of the Visual Cortex. Individual neurons respond to stimuli only in a restricted region of the visual field known as the Receptive Field. A collection of such fields overlap to cover the entire visual area.<ref>Definition - What Does Convolutional Neural Network (CNN) Mean? [https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53 Towards Data Science]</ref>

← Older revision		Revision as of 17:58, 4 September 2023
Line 1:		Line 1:
−	A '''Convolutional Neural Network (~~ConvNet/~~CNN)''' is a ~~Deep Learning algorithm which can take in an input~~ image~~, assign importance (learnable weights~~ and ~~biases)~~ to ~~various aspects/objects in the image and be able~~ to ~~differentiate one from the~~ other. The ~~pre-processing required~~ in ~~a ConvNet is much lower as compared to other classification algorithms~~. ~~While in primitive methods filters are hand-engineered~~, ~~with enough training~~, ~~ConvNets have~~ the ~~ability to learn these filters/characteristics~~. The ~~architecture~~ of a ~~ConvNet is analogous~~ to ~~that~~ of the ~~connectivity pattern~~ of ~~Neurons in~~ the ~~Human Brain~~ and ~~was inspired~~ by the ~~organization of~~ the ~~Visual Cortex~~. ~~Individual neurons respond~~ to ~~stimuli only~~ in ~~a restricted region~~ of ~~the visual field known~~ as the ~~Receptive Field. A collection~~ of ~~such fields overlap to cover~~ the ~~entire visual area~~.~~<ref>Definition - What Does Convolutional Neural Network~~ (~~CNN~~) ~~Mean? [https~~:~~//towardsdatascience~~.~~com/~~a~~-comprehensive-guide-~~to-~~convolutional~~-neural~~-networks-the-eli5-way-3bd2b1164a53 Towards Data Science~~]~~</ref>~~	+	A '''Convolutional Neural Network (CNN)''' is a class of deep neural networks most commonly applied to visual tasks, such as image and video recognition. CNNs are also known for their exceptional performance in tasks related to spatial hierarchies, making them applicable to other forms of data with similar properties.
		+
		+
		+	== History ==
		+	The development of CNNs can be traced back to the Neocognitron introduced by Kunihiko Fukushima in 1980. However, CNNs gained popularity after AlexNet, a particular kind of CNN architecture, won the ImageNet Large Scale Visual Recognition Challenge (ILSVRC) in 2012.
		+
		+
		+	== Architecture ==
		+	*Convolutional Layer: The convolutional layer is the core building block of a CNN. It filters inputs for useful information. These convolutional layers have parameters that are learned to identify various types of features like edges, corners, and textures.
		+	*Pooling Layer: Pooling (or subsampling) reduces the dimensionality of each feature map and retains the most essential information, thus making the detection of features invariant to scale and orientation changes.
		+	*Fully Connected Layer: Fully connected layers perform classification on the features formed by the convolutional layers and down-sampled by the pooling layers. In a CNN, the fully connected layer is generally followed by an output layer, activation function, or classifier.
		+	*Activation Functions: Activation functions introduce nonlinear properties into the network. The commonly used activation functions in CNNs are ReLU (Rectified Linear Unit), Sigmoid, and Tanh.
		+
		+
		+	== Applications ==
		+	*Image Recognition: CNNs are extensively used in image classification tasks, for everything from identifying objects in photos to diagnosing diseases from medical imaging.
		+	*Natural Language Processing: Although less common, CNNs have been successfully applied in Natural Language Processing (NLP), often combined with other types of neural networks like RNNs.
		+	*Video Analysis: CNNs are also employed in video analysis for tasks such as object detection and recognition over time, activity recognition, and video classification.
		+
		+
		+	== Training ==
		+	*Backpropagation: CNNs, like all neural networks, use backpropagation for training. It involves computing the gradient of the loss function concerning each weight by the chain rule.
		+	*Optimization Algorithms: Common optimization algorithms used for training CNNs include Stochastic Gradient Descent (SGD), Adam, and RMSprop.
		+	*Software Libraries: Popular software libraries for implementing CNNs include TensorFlow, Keras, and PyTorch.
		+
		+
		+	== Challenges and Limitations ==
		+	*Requires a large amount of labeled data for training
		+	*Computationally expensive
		+	*Vulnerable to adversarial attacks
		+
		+
		+	== Future Trends ==
		+	*Real-time processing
		+	*Energy-efficient CNN architectures
		+	*Integration with other neural network architectures for various applications
		+
		+
		+	== See Also ==
		+	*[[Deep Learning]]
		+	*[[Artificial Neural Network (ANN)]]
		+	*[[Natural Language Processing (NLP)]]
		+	*[[Neural Network]]
		+	__NOTOC__

Convolutional Neural Network (CNN) - Revision history

User at 17:58, 4 September 2023

User: Created page with "A '''Convolutional Neural Network (ConvNet/CNN)''' is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various a..."