Convolutional neural network
A convolutional neural network is a neural-network architecture that applies learned convolutional operations to extract patterns from structured data.
Natural Language Processing in Action, Second Edition
A convolutional neural network is a neural network that applies learned convolutional filters to structured inputs to detect local patterns and construct hierarchical representations.
- How it is used
- Uses learned sequence filters to support NLP tasks including text classification and word tagging.
- How deep it goes
- Builds from a sliding-stencil explanation and hand-set filters to PyTorch layers, tensor shapes, embeddings, pooling and training.
- Examples
- Works through an adverb–verb–noun detector and a disaster-tweet classifier, with exercises on Morse code and sarcasm detection.
- What you can do after
- Build it from scratch with PyTorch `Conv1d`, aligned embedding channels, pooling and fixed-length inputs for text classification.
Sutskever's List
A convolutional neural network is a neural network architecture that applies learned convolutional filters to structured input such as images.
- How it is used
- Frames CNNs as the architecture behind AlexNet’s breakthrough in image classification and a contrast to later sequence models.
- How deep it goes
- Explains CNNs through competition results, end-to-end learning, and local interactions, without detailing filter implementation.
- Examples
- Compares AlexNet’s 15% top-5 error with conventional methods’ roughly 26%, then traces GoogLeNet and R-CNN.
- What you can do after
- Evaluate/understand why AlexNet displaced handcrafted image classifiers and why local interactions constrained convolutional sequence models.
Read how Sutskever's List explains convolutional neural network →
Deep Learning with R, Third Edition
A convolutional neural network is a neural network that applies convolution-based operations to learn spatial or locally structured features, especially from grid-like data such as images.
- How it is used
- Applies convnets to image classification, from handwritten digits to dogs-versus-cats images.
- How deep it goes
- Works through model construction, training and evaluation, then explains local convolutions and feature maps without a mathematical derivation.
- Examples
- Builds a MNIST digit classifier and a larger dogs-versus-cats classifier with R/Keras model listings.
- What you can do after
- Build it from scratch with R/Keras for image classification and evaluate when small-data overfitting calls for other strategies.
Read how Deep Learning with R, Third Edition explains convolutional neural network →
Deep Learning with Python, Third Edition
A neural network built from convolution and pooling layers that processes data with local spatial structure.
- How it is used
- Uses local spatial patterns to classify images and compares that approach with densely connected networks.
- How deep it goes
- Works through Keras models and explains patches, feature maps, and pattern hierarchies without a heavy mathematical derivation.
- Examples
- Builds MNIST and dogs-versus-cats classifiers, then tests a 1D ConvNet on temperature forecasting.
- What you can do after
- Build it from scratch in Keras for image classification, or apply feature extraction and fine-tuning to a pretrained ConvNet.
Read how Deep Learning with Python, Third Edition explains convolutional neural network →
Evolutionary Deep Learning: Genetic algorithms and neural networks
A convolutional neural network is a neural network that processes structured inputs with learned convolutional filters, typically for tasks such as image classification.
- How it is used
- Supports image classification by extracting localized features and provides an architecture target for evolutionary optimization.
- How deep it goes
- Walks through Keras construction, parameter summaries, training behavior, pooling effects, and gene-based architecture encoding.
- Examples
- Uses Fashion-MNIST, a three-convolution-layer Keras model, a oversized single-layer model, and variable-length gene sequences.
- What you can do after
- Build from scratch a Keras CNN and encode its layer architecture for genetic optimization.
Deep Learning with PyTorch, Second Edition
A neural network that uses convolutional layers to process spatial data.
- How it is used
- Uses convolutions to make image classifiers less dependent on dense connections and an object's position.
- How deep it goes
- Works through learned kernels, pooling, layer shapes and a PyTorch classifier, including a missing-reshape error.
- Examples
- Uses blurred and edge-highlighted bird images, a bird-versus-airplane classifier and a 3D tumor-detection design.
- What you can do after
- Apply it via PyTorch nn.Conv2d and nn.MaxPool2d to assemble a bird-versus-airplane classifier.
Read how Deep Learning with PyTorch, Second Edition explains convolutional neural network →
Distributed Machine Learning Patterns
A convolutional neural network is a neural network model that uses convolutional operations to learn spatial or locally structured representations.
- How it is used
- Uses CNNs as candidate image classifiers in local and distributed training pipelines.
- How deep it goes
- Works through TensorFlow layer construction, compilation, callbacks, variants, and distributed execution; defers convolution operation internals.
- Examples
- Uses LeNet for 10-entity video classification and Fashion-MNIST CNN variants trained through parallel TFJobs.
- What you can do after
- Operate/configure it as a TensorFlow CNN, train variants, and run their distributed selection-and-serving workflow.
Read how Distributed Machine Learning Patterns explains convolutional neural network →
Inside Deep Learning: Math, Algorithms, Models
A neural-network design that uses convolutional layers to exploit spatial structure in image data.
- How it is used
- Uses spatial priors to build and train image-classification models.
- How deep it goes
- Explains image tensors, convolution mechanics, pooling tradeoffs, and PyTorch construction, but defers advanced large-image techniques.
- Examples
- Demonstrates MNIST tensors, RGB channels, one-dimensional filters, max-pooling architectures, and exploratory CIFAR10 experiments.
- What you can do after
- Builds and evaluates PyTorch CNNs while testing layers, pooling, filters, and image transforms.
Read how Inside Deep Learning: Math, Algorithms, Models explains convolutional neural network →
Graph Neural Networks in Action
A convolutional neural network is a neural network that applies learned convolutional filters to exploit local spatial structure, especially in grid-like data such as images.
- How it is used
- Applies learned filters to fixed pixel grids for computer-vision tasks such as segmentation and classification.
- How deep it goes
- Explains the filter–activation–pooling pipeline and contrasts grid convolution with graph convolution.
- Examples
- Uses an animal image and a sliding kernel to illustrate feature extraction, alongside image segmentation and classification.
- What you can do after
- Evaluate CNN layers as sequences of learned filtering, nonlinear activation, and spatial reduction.
Read how Graph Neural Networks in Action explains convolutional neural network →
Learn Generative AI with PyTorch
A neural network architecture using convolution operations, here applied to multidimensional music data to extract spatial features.
- How it is used
- Uses local feature extraction and spatial hierarchy to process images and support high-resolution image-generation models.
- How deep it goes
- Explains local connectivity, shared weights, filters, kernel size, stride, padding, upsampling, and batch normalization without detailed derivations.
- Examples
- Demonstrates anime-face generation, high-resolution color images, and human-face reconstruction with an eyeglasses dataset.
- What you can do after
- Build it from scratch into image-generation systems using convolutional and transposed convolutional layers.
Read how Learn Generative AI with PyTorch explains convolutional neural network →
The AI Pocket Book
A neural network architecture that applies convolutions and downsampling to learn spatial features from grid-like data such as images.
- How it is used
- Processes images and other media, chiefly for categorization and multimodal systems.
- How deep it goes
- Explains learned filters, repeated convolutions, downsampling, and spatial abstraction without deriving the mathematics.
- Examples
- Uses strawberry classification, cat parts, learned visual filters, pixel downsampling, and diffusion-based image repair.
- What you can do after
- Evaluate/understand how CNNs transform pixels into increasingly abstract features for image classification and multimodal generation.
Read how The AI Pocket Book explains convolutional neural network →
Data Without Labels
A convolutional neural network is a neural-network model that uses convolutional operations to learn spatial or locally structured features.
- How it is used
- Deploys learned spatial features for image, video, speech-recognition, and computer-vision solutions.
- How deep it goes
- Explains tensors, filters, feature maps, activations, pooling, classification, and shared parameters, but provides only a brief overview.
- Examples
- Uses cat-versus-dog classification, license-plate reading, cancer-cell detection, bone-fracture detection, and facial recognition.
- What you can do after
- Apply convolutional architectures to image classification, image processing, object detection, and related recognition problems.
Read how Data Without Labels explains convolutional neural network →
Math and Architectures of Deep Learning
A neural network that uses convolutional layers to process structured grid data such as images.
- How it is used
- Processes image grids for image classification and object detection by extracting reusable local visual patterns.
- How deep it goes
- Explains local connections, shared weights, hierarchical features, pooling, downsampling, and output-size effects.
- Examples
- Uses edge, corner, ear, eye, nose, and face detectors, plus 2 × 2 max- and average-pooling examples.
- What you can do after
- Understand how local filters and pooling build image representations for classification and object detection.
Read how Math and Architectures of Deep Learning explains convolutional neural network →
Grokking Machine Learning
A neural-network architecture that uses convolutional operations to learn localized and hierarchical patterns, especially in structured data such as images.
- How it is used
- Deploys convolutional architectures for image recognition and semantic segmentation, especially when fully connected networks lose spatial information.
- How deep it goes
- Explains the sliding-window convolution intuition but stops short of a complete implementation or formal derivation.
- Examples
- Uses MNIST, CIFAR-10, CIFAR-100, complicated pictures, and faces as image examples.
- What you can do after
- Evaluate when convolutional architectures better preserve useful image information than fully connected networks.
Read how Grokking Machine Learning explains convolutional neural network →
Grokking AI Algorithms, Second Edition
A convolutional neural network is a neural-network architecture that applies learned convolutional filters and typically pooling operations to detect spatial patterns in structured data such as images.
- How it is used
- Applies local filters and pooling to recognize spatial patterns in images and related structured data.
- How deep it goes
- Explains local convolution, pooling, feature hierarchies, and spatial structure without detailing filter-learning mathematics or implementation code.
- Examples
- Uses pixels, neighboring regions, edges, faces, cars, X-rays, body scans, and local word patterns as examples.
- What you can do after
- Evaluate when spatial feature detection suits image classification, OCR, medical scans, or local-pattern text classification.
Read how Grokking AI Algorithms, Second Edition explains convolutional neural network →
Delivering Machine Learning Projects: From design to deployment
A convolutional network is a neural network architecture that applies local filters over structured inputs to capture spatial or local patterns.
- How it is used
- Uses local spatial structure to support computer-vision and image-recognition tasks.
- How deep it goes
- Explains local filtering, sliding windows, normalization, and pooling, but stops short of mathematical or implementation detail.
- Examples
- Contrasts pixel-by-pixel perceptron processing with learned filters sliding across neighboring image pixels.
- What you can do after
- Evaluate why local filtering is appropriate for image-recognition problems.
Machine Learning Algorithms in Depth
A convolutional neural network is a neural-network architecture that uses convolution and pooling operations to process structured spatial or grid-like data.
- How it is used
- Uses learned spatial features for image classification and image-search applications.
- How deep it goes
- Derives convolution and pooling operations, then applies them in a LeNet workflow; broader architecture internals are not developed.
- Examples
- Demonstrates LeNet on MNIST handwritten digits and names ResNet50 image search as an advanced application.
- What you can do after
- Apply CNNs via Keras/TensorFlow to train, evaluate, and visualize an MNIST classifier.
Read how Machine Learning Algorithms in Depth explains convolutional neural network →
Inside AI
A convolutional neural network is a neural-network model that applies learned convolutional filters to structured data such as images.
- How it is used
- Applies deep learning to classify structured image data in computer-vision projects.
- How deep it goes
- Explains image encoding, training, and testing, but does not detail convolution operations or network internals.
- Examples
- Uses RGB photographs of cats and dogs, labeled training data, and unseen test images.
- What you can do after
- Evaluate image-classification performance by separating training data from previously unseen test data.
Related concepts
- language modelkind of · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
- convolutional layeruses · A convolutional layer is a neural-network layer that applies shared learned filters to local regions of structured input to produce feature representations.
- poolinguses · Pooling is an operation that aggregates values or representations across regions or axes to produce a reduced or summarized representation.
- convolution kerneluses · A convolution kernel is a small array or tensor of weights applied across local input regions to compute weighted responses and detect or transform patterns.
- U-Netincludes · U-Net is an encoder-decoder neural-network architecture with downsampling and upsampling paths connected by skip connections for combining features at multiple spatial scales.
- LeNetincludes · LeNet is an early convolutional neural-network architecture designed primarily for handwritten-digit and character recognition.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.