# Improving Efficiency in Convolutional Neural Network with Multilinear Filters

arXiv: https://arxiv.org/abs/1709.09902 · HTML: https://mlsanity.com/paper/1709.09902/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Sep 2017.

## Built on by (most cited first, 8 of 8)

- [Single-Layer Vision Transformers for More Accurate Early Exits with Less Overhead](https://mlsanity.com/paper/2105.09121/) (2021)
- [Structured Pruning Adapters](https://mlsanity.com/paper/2211.10155/) (2022)
- [Efficient High-Resolution Deep Learning: A Survey](https://mlsanity.com/paper/2207.13050/) (2022)
- [Augmented Bilinear Network for Incremental Multi-Stock Time-Series Classification](https://mlsanity.com/paper/2207.11577/) (2022)
- [Efficient Visual Recognition with Deep Neural Networks: A Survey on Recent Advances and New Directions](https://mlsanity.com/paper/2108.13055/) (2021)
- [Knowledge Distillation By Sparse Representation Matching](https://mlsanity.com/paper/2103.17012/) (2021)
- [Efficient Online Processing with Deep Neural Networks](https://mlsanity.com/paper/2306.13474/) (2023)
- [Convolutional Neural Network Compression via Dynamic Parameter Rank Pruning](https://mlsanity.com/paper/2401.08014/) (2024)

## Abstract (arXiv)

The excellent performance of deep neural networks has enabled us to solve several automatization problems, opening an era of autonomous devices. However, current deep net architectures are heavy with millions of parameters and require billions of floating point operations. Several works have been developed to compress a pre-trained deep network to reduce memory footprint and, possibly, computation. Instead of compressing a pre-trained network, in this work, we propose a generic neural network layer structure employing multilinear projection as the primary feature extractor. The proposed architecture requires several times less memory as compared to the traditional Convolutional Neural Networks (CNN), while inherits the similar design principles of a CNN. In addition, the proposed architecture is equipped with two computation schemes that enable computation reduction or scalability. Experimental results show the effectiveness of our compact projection that outperforms traditional CNN, while requiring far fewer parameters.