# Fast Training of Convolutional Networks through FFTs

arXiv: https://arxiv.org/abs/1312.5851 · HTML: https://mlsanity.com/paper/1312.5851/

## Key facts

- Cited by 89 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Dec 2013.

## Built on by (most cited first, 30 of 89)

- [FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting](https://mlsanity.com/paper/2201.12740/) (2022)
- [Neural Operator: Learning Maps Between Function Spaces](https://mlsanity.com/paper/2108.08481/) (2021)
- [FNet: Mixing Tokens with Fourier Transforms](https://mlsanity.com/paper/2105.03824/) (2021)
- [Pruning and Quantization for Deep Neural Network Acceleration: A Survey](https://mlsanity.com/paper/2101.09671/) (2021)
- [Scaling Local Self-Attention for Parameter Efficient Visual Backbones](https://mlsanity.com/paper/2103.12731/) (2021)
- [Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs](https://mlsanity.com/paper/2203.06717/) (2022)
- [ETSformer: Exponential Smoothing Transformers for Time-series Forecasting](https://mlsanity.com/paper/2202.01381/) (2022)
- [A Mathematical Guide to Operator Learning](https://mlsanity.com/paper/2312.14688/) (2023)
- [Deep Frequency Filtering for Domain Generalization](https://mlsanity.com/paper/2203.12198/) (2022)
- [RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition](https://mlsanity.com/paper/2105.01883/) (2021)
- [AutoFL: Enabling Heterogeneity-Aware Energy Efficient Federated Learning](https://mlsanity.com/paper/2107.08147/) (2021)
- [DCTdiff: Intriguing Properties of Image Generative Modeling in the DCT Space](https://mlsanity.com/paper/2412.15032/) (2024)
- [RepMLPNet: Hierarchical Vision MLP with Re-parameterized Locality](https://mlsanity.com/paper/2112.11081/) (2021)
- [Conv-Basis: A New Paradigm for Efficient Attention Inference and Gradient Computation in Transformers](https://mlsanity.com/paper/2405.05219/) (2024)
- [Frequency Domain-based Dataset Distillation](https://mlsanity.com/paper/2311.08819/) (2023)
- [projUNN: efficient method for training deep networks with unitary matrices](https://mlsanity.com/paper/2203.05483/) (2022)
- [Explaining the physics of transfer learning a data-driven subgrid-scale closure to a different turbulent flow](https://mlsanity.com/paper/2206.03198/) (2022)
- [Transform Once: Efficient Operator Learning in Frequency Domain](https://mlsanity.com/paper/2211.14453/) (2022)
- [Frequency Dynamic Convolution for Dense Image Prediction](https://mlsanity.com/paper/2503.18783/) (2025)
- [Truly Scale-Equivariant Deep Nets with Fourier Layers](https://mlsanity.com/paper/2311.02922/) (2023)
- [Is Complexity Required for Neural Network Pruning? A Case Study on Global Magnitude Pruning](https://mlsanity.com/paper/2209.14624/) (2022)
- [Spectral Leakage and Rethinking the Kernel Size in CNNs](https://mlsanity.com/paper/2101.10143/) (2021)
- [Accelerating Neural ODEs Using Model Order Reduction](https://mlsanity.com/paper/2105.14070/) (2021)
- [Manipulating Identical Filter Redundancy for Efficient Pruning on Deep and Complicated CNN](https://mlsanity.com/paper/2107.14444/) (2021)
- [Resource-Efficient Deep Learning: A Survey on Model-, Arithmetic-, and Implementation-Level Techniques](https://mlsanity.com/paper/2112.15131/) (2021)
- [Bitwidth-Adaptive Quantization-Aware Neural Network Training: A Meta-Learning Approach](https://mlsanity.com/paper/2207.10188/) (2022)
- [As large as it gets: Learning infinitely large Filters via Neural Implicit Functions in the Fourier Domain](https://mlsanity.com/paper/2307.10001/) (2023)
- [k-strip: A novel segmentation algorithm in k-space for the application of skull stripping](https://mlsanity.com/paper/2205.09706/) (2022)
- [Adaptive Sampling of k-Space in Magnetic Resonance for Rapid Pathology Prediction](https://mlsanity.com/paper/2406.04318/) (2024)
- [Distributed-Order Fractional Graph Operating Network](https://mlsanity.com/paper/2411.05274/) (2024)

## Abstract (arXiv)

Convolutional networks are one of the most widely employed architectures in computer vision and machine learning. In order to leverage their ability to learn complex functions, large amounts of data are required for training. Training a large convolutional network to produce state-of-the-art results can take weeks, even when using modern GPUs. Producing labels using a trained network can also be costly when dealing with web-scale datasets. In this work, we present a simple algorithm which accelerates training and inference by a significant factor, and can yield improvements of over an order of magnitude compared to existing state-of-the-art implementations. This is done by computing convolutions as pointwise products in the Fourier domain while reusing the same transformed feature map many times. The algorithm is implemented on a GPU architecture and addresses a number of related challenges.