Harmonic Networks: Integrating Spectral Information into CNNs
arXiv:1812.03205
Abstract
Convolutional neural networks (CNNs) learn filters in order to capture local correlation patterns in feature space. In contrast, in this paper we propose harmonic blocks that produce features by learning optimal combinations of spectral filters defined by the Discrete Cosine Transform. The harmonic blocks are used to replace conventional convolutional layers to construct partial or fully harmonic CNNs. We extensively validate our approach and show that the introduction of harmonic blocks into state-of-the-art CNN baseline architectures results in comparable or better performance in classification tasks on small NORB, CIFAR10 and CIFAR100 datasets.
References in corpus (12)
- Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
- Wide Residual Networks
- Predicting Parameters in Deep Learning
- Compressing Neural Networks with the Hashing Trick
- Speeding up Convolutional Neural Networks with Low Rank Expansions
- Towards Image Understanding from Deep Compression without Decoding
- Wavelet Convolutional Neural Networks for Texture Classification
- Harmonic Networks: Deep Translation and Rotation Equivariance
- Compressing the Input for CNNs with the First-Order Scattering Transform
- Scaling the Scattering Transform: Deep Hybrid Networks
- Deep Clustered Convolutional Kernels
- Wavelet based edge feature enhancement for convolutional neural networks