/SEEK-A-Framework-of-Superpixel-Learning-with-CNN-Features-for-Unsupervised-Segmentation

Primary LanguageJupyter NotebookCreative Commons Attribution 4.0 InternationalCC-BY-4.0

CC BY 4.0 PyTorch

alt text

SEEK: A Framework of Superpixel Learning with CNN Features for Unsupervised Segmentation

This repo explains the working of our Unsupervised Segemntaion Algorithm published in Electronics (MDPI) journal.

You can access the full paper here.

Abstract

Supervised semantic segmentation algorithms have been a hot area of exploration recently, but now the attention is being drawn towards completely unsupervised semantic segmentation. In an unsupervised framework, neither the targets nor the ground truth labels are provided to the network. That being said, the network is unaware about any class instance or object present in the given data sample. So, we propose a convolutional neural network (CNN) based architecture for unsupervised segmentation. We used the squeeze and excitation network, due to its peculiar ability to capture the features’ interdependencies, which increases the network’s sensitivity to more salient features. We iteratively enable our CNN architecture to learn the target generated by a graph-based segmentation method, while simultaneously preventing our network from falling into the pit of over-segmentation. Along with this CNN architecture, image enhancement and refinement techniques are exploited to improve the segmentation results. Our proposed algorithm produces improved segmented regions that meet the human level segmentation results. In addition, we evaluate our approach using different metrics to show the quantitative outperformance.

Network Architecture

Following images shows Complete Network Architecture: Here, contrast and texture enhancement (CTE)-Block represents the contrast and texture enhancement block, CE Loss is the cross-entropy loss, complete architecture of SE-Block is explained in Figure 2, black arrows show the forward pass, and blue arrows show the backward error propagation. For details on architecture kindly visit here .

alt text

Algorithm 1 (Unsupervised Semantic Segmentation)

alt text

Algorithm 2 (K nearest neighbour)

This flow chart shows how we find the value of K for K-means clustering algorithm, implementation is provided in code. alt text

Example Output

Segmentation results. From top to bottom: (a) Original images, (b) segmentation results of proposed algorithm.

alt text

Citation

If you like to use our work please consider citing us;

@article{ilyas2020seek,
  title={SEEK: A Framework of Superpixel Learning with CNN Features for Unsupervised Segmentation},
  author={Ilyas, Talha and Khan, Abbas and Umraiz, Muhammad and Kim, Hyongsuk},
  journal={Electronics},
  volume={9},
  number={3},
  pages={383},
  year={2020},
  publisher={Multidisciplinary Digital Publishing Institute}

OR via google scholar

This work is licensed under a Creative Commons Attribution 4.0 International License.

CC BY 4.0