Transposed convolution
A transposed convolution is a learnable convolution-like operation corresponding to the transpose of a convolution operator and often used to increase spatial resolution.
Learn Generative AI with PyTorch
The 2D computation performed by a transposed convolutional layer to expand an image or feature map.
- How it is used
- Upsamples feature maps in image generators and decoders, with a further use extending music-generation vectors across bars.
- How deep it goes
- Works through a hand-weighted 2×2 calculation and output shape, then shows a multilayer PyTorch generator.
- Examples
- Uses a 2×2 numerical image and a DCGAN generator for RGB anime faces.
- What you can do after
- Apply it via PyTorch nn.ConvTranspose2d to expand feature maps and assemble an image generator.
Read how Learn Generative AI with PyTorch explains transposed convolution →
Inside Deep Learning: Math, Algorithms, Models
A convolution-like operation used to increase spatial resolution in a tensor, often to undo downsampling from pooling in dense prediction models.
- How it is used
- Expands pooled feature maps for pixel-level segmentation, U-Net decoding, and convolutional image generators.
- How deep it goes
- Explains overlapping contributions, borders, stride, PyTorch parameters, architectures, and artifacts without deriving gradients.
- Examples
- Uses 2×2 mapping diagrams, a pooled segmentation model, U-Net blocks, autoencoder exercises, and a two-stage GAN generator.
- What you can do after
- Builds PyTorch segmentation or generator layers that restore spatial size after pooling and evaluates artifact-reduction alternatives.
Read how Inside Deep Learning: Math, Algorithms, Models explains transposed convolution →
Build a Text-to-Image Generator (from Scratch)
A transposed convolutional layer increases spatial resolution by applying a learned convolution-like operation that can expand spatial dimensions.
- How it is used
- Upsamples feature maps in U-Net expansive paths and reconstructs high-resolution images.
- How deep it goes
- Explains output growth, stride control, kernel application, and a worked 2D calculation, then implements the layer in PyTorch.
- Examples
- Uses a hand-weighted 2 × 2 image example and a PyTorch UnetUp implementation.
- What you can do after
- Build it from scratch by configuring ConvTranspose2d and placing it inside a U-Net up block.
Read how Build a Text-to-Image Generator (from Scratch) explains transposed convolution →
Math and Architectures of Deep Learning
A convolution-like operation often described via the transpose of the convolution matrix, used here as a motivation for understanding the matrix form of convolution.
- How it is used
- Uses transposed convolution to upsample feature maps and reconstruct inputs from embeddings in decoder networks.
- How deep it goes
- Derives the matrix intuition, output-size formula, tensor shapes, and PyTorch use, but defers autoencoder training details.
- Examples
- Demonstrates one-dimensional matrix transposition, a 3-to-5 output calculation, and 2D PyTorch upsampling from 2 × 2 to 4 × 4.
- What you can do after
- Apply transposed convolution via PyTorch ConvTranspose2d and calculate its output dimensions from stride, kernel, and padding.
Read how Math and Architectures of Deep Learning explains transposed convolution →
Grokking AI Algorithms, Second Edition
A transposed convolution is a learnable convolutional operation that projects a lower-resolution representation into a higher-resolution output space.
- How it is used
- Uses transposed convolution in U-Net decoder paths to enlarge compressed feature maps during generative image reconstruction.
- How deep it goes
- Explains filter-based patch painting, overlap accumulation, learned weights, and a small implementation while stopping short of deeper theory.
- Examples
- Uses nearest-neighbor duplication, an artist reconstruction analogy, and a 4×4-to-8×8 U-Net feature-map example.
- What you can do after
- Builds a basic upsampling layer using filters, stride, padding, output dimensions, and accumulated overlapping contributions.
Read how Grokking AI Algorithms, Second Edition explains transposed convolution →
Related concepts
- up blockused by · An up block is a U-Net component that increases spatial resolution through operations such as transposed convolution.
- upsamplingimplements · Upsampling is a process that increases the number, density, or spatial or temporal resolution of samples or values in a dataset or representation by adding, duplicating, or estimating them.
- 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.
- convolutional layercontrasts with · A convolutional layer is a neural-network layer that applies shared learned filters to local regions of structured input to produce feature representations.
- decodeused by · Decoding is the process of transforming an encoded or internal representation into an interpretable or target-form output.
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.