Improved image clustering with deep semantic embedding

Guo, Jun; Xuan, Yue; Xu, Pengfei; Bai, Hao; Liu, Baoying

doi:10.1016/j.patrec.2018.10.022

Cited by 9 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Guo et al in their studies used an auto-encoder with the multilayer-based deep learning to perform the semantic feature embedding and dimensional reduction. The experimental results show that proposed approaches get overwhelming performance over numerous clustering methods in the literature (Guo et al, 2020). In the dimensional reduction area, some relevant and important information normally can be lost.…”

Section: Literature Reviewmentioning

confidence: 99%

Efficient path planning of drone swarms over clustered human crowds in social events

Bulut¹,

Bektaş

Yavuz

2021

IJIUS

View full text Add to dashboard Cite

PurposeIn this study, supervision and control of the possible problems among people over a large area with a limited number of drone cameras and security staff is established.Design/methodology/approachThese drones, namely unmanned aerial vehicles (UAVs) will be adaptively and automatically distributed over the crowds to control and track the communities by the proposed system. Since crowds are mobile, the design of the drone clusters will be simultaneously re-organized according to densities and distributions of people. An adaptive and dynamic distribution and routing mechanism of UAV fleets for crowds is implemented to control a specific given region. The nine popular clustering algorithms have been used and tested in the presented mechanism to gain better performance.FindingsThe nine popular clustering algorithms have been used and tested in the presented mechanism to gain better performance. An outperformed clustering performance from the aggregated model has been received when compared with a singular clustering method over five different test cases about crowds of human distributions. This study has three basic components. The first one is to divide the human crowds into clusters. The second one is to determine an optimum route of UAVs over clusters. The last one is to direct the most appropriate security personnel to the events that occurred.Originality/valueThis study has three basic components. The first one is to divide the human crowds into clusters. The second one is to determine an optimum route of UAVs over clusters. The last one is to direct the most appropriate security personnel to the events that occurred.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Efficient path planning of drone swarms over clustered human crowds in social events

Bulut¹,

Bektaş

Yavuz

2021

IJIUS

View full text Add to dashboard Cite

show abstract

“…Pseudo graphs and pseudo labels can benefit from the uncertain knowledge gained from online training, and are further used to monitor similar learning. Image clustering with deep semantic embedding (DSEC) [ 15 ] extracts the total semantic (attribute) features from the data set firstly, and then employs a deep semantic embedding auto-encoder to refine the lower dimensional multi-features representation. The final clustering work is implemented by iteratively optimizing a KL divergence-based clustering objective.…”

Section: Related Workmentioning

confidence: 99%

Image Clustering Algorithm Based on Predefined Evenly-Distributed Class Centroids and Composite Cosine Distance

Zhu

Wang

2022

Entropy

View full text Add to dashboard Cite

The clustering algorithms based on deep neural network perform clustering by obtaining the optimal feature representation. However, in the face of complex natural images, the cluster accuracy of existing clustering algorithms is still relatively low. This paper presents an image clustering algorithm based on predefined evenly-distributed class centroids (PEDCC) and composite cosine distance. Compared with the current popular auto-encoder structure, we design an encoder-only network structure with normalized latent features, and two effective loss functions in latent feature space by replacing the Euclidean distance with a composite cosine distance. We find that (1) contrastive learning plays a key role in the clustering algorithm and greatly improves the quality of learning latent features; (2) compared with the Euclidean distance, the composite cosine distance can be more suitable for the normalized latent features and PEDCC-based Maximum Mean Discrepancy (MMD) loss function; and (3) for complex natural images, a self-supervised pretrained model can be used to effectively improve clustering performance. Several experiments have been carried out on six common data sets, MNIST, Fashion-MNIST, COIL20, CIFAR-10, STL-10 and ImageNet-10. Experimental results show that our method achieves the best clustering effect compared with other latest clustering algorithms.

show abstract

“…As a fundamental technique in data mining [2,3], clustering analysis aims at dividing data objects into several groups such that data objects in each group are similar to one another and dissimilar to data objects in different groups [4,5]. Over the years, clustering algorithms are widely used in data analysis in different domains, such as text data [6,7], customer data [8,9], image data [10,11] and medical data [12,13].…”

Section: Introductionmentioning

confidence: 99%

Clustering Algorithm Based on Sparse Feature Vector without Specifying Parameter

2020

Teh. vjesn.

View full text Add to dashboard Cite

Improved image clustering with deep semantic embedding

Cited by 9 publications

References 11 publications

Efficient path planning of drone swarms over clustered human crowds in social events

Efficient path planning of drone swarms over clustered human crowds in social events

Image Clustering Algorithm Based on Predefined Evenly-Distributed Class Centroids and Composite Cosine Distance

Clustering Algorithm Based on Sparse Feature Vector without Specifying Parameter

Contact Info

Product

Resources

About