# CUNet: A Compact Unsupervised Network for Image Classification

arXiv: https://arxiv.org/abs/1607.01577 · HTML: https://mlsanity.com/paper/1607.01577/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Jul 2016.

## Built on by (most cited first, 4 of 4)

- [Discriminative Semantic Feature Pyramid Network with Guided Anchoring for Logo Detection](https://mlsanity.com/paper/2108.13775/) (2021)
- [PDFNet: Pointwise Dense Flow Network for Urban-Scene Segmentation](https://mlsanity.com/paper/2109.10083/) (2021)
- [SimpleFusion: A Simple Fusion Framework for Infrared and Visible Images](https://mlsanity.com/paper/2406.19055/) (2024)
- [SASE: A Searching Architecture for Squeeze and Excitation Operations](https://mlsanity.com/paper/2411.08333/) (2024)

## Abstract (arXiv)

In this paper, we propose a compact network called CUNet (compact unsupervised network) to counter the image classification challenge. Different from the traditional convolutional neural networks learning filters by the time-consuming stochastic gradient descent, CUNet learns the filter bank from diverse image patches with the simple K-means, which significantly avoids the requirement of scarce labeled training images, reduces the training consumption, and maintains the high discriminative ability. Besides, we propose a new pooling method named weighted pooling considering the different weight values of adjacent neurons, which helps to improve the robustness to small image distortions. In the output layer, CUNet integrates the feature maps gained in the last hidden layer, and straightforwardly computes histograms in non-overlapped blocks. To reduce feature redundancy, we implement the max-pooling operation on adjacent blocks to select the most competitive features. Comprehensive experiments are conducted to demonstrate the state-of-the-art classification performances with CUNet on CIFAR-10, STL-10, MNIST and Caltech101 benchmark datasets.