Xiaokun Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiaokun Zhang has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Xin Jin
Department of Cyber Security, Beijing Electronic Science and Technology Institute, Beijing, China
4 shared publications - Xuying Zhao
Beijing Electronic Science and Technology Institute
3 shared publications Geng Zhao
Beijing Electronic Science and Technology Institute
2 shared publicationsKejun Zhang
Beijing Electronic Science and Technology Institute, China
2 shared publications- Xiaodong Li
School of Foreign Studies, University of Science and Technology Beijing, Beijing, China
2 shared publications - Binbin Lv1 shared publication
Latest Publications
- 19 September 2024
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- Vol. 16 (3) , 1691-1700
Image recoloring for multiple types of Color Vision Deficiency
- 19 September 2024
- Published by Springer Nature in International Journal of Machine Learning and Cybernetics
- Vol. 16 (3) , 1691-1700
Image recoloring for multiple types of Color Vision Deficiency
- Open Access
TBFormer: Two-Branch Transformer for Image Forgery Localization
Image forgery localization aims to identify forged regions by capturing subtle traces from high-quality discriminative features. In this paper, we propose a Transformer-style network with two feature extraction branches for image forgery localization, and it is named as Two-Branch Transformer (TBFormer). Firstly, two feature extraction branches are elaborately designed, taking advantage of the discriminative stacked Transformer layers, for both RGB and noise domain features. Secondly, an Attention-aware Hierarchical-feature Fusion Module (AHFM) is proposed to effectively fuse hierarchical features from two different domains. Although the two feature extraction branches have the same architecture, their features have significant differences since they are extracted from different domains. We adopt position attention to embed them into a unified feature domain for hierarchical feature investigation. Finally, a Transformer decoder is constructed for feature reconstruction to generate the predicted mask. Extensive experiments on publicly available datasets demonstrate the effectiveness of the proposed model.
Open AccessTBFormer: Two-Branch Transformer for Image Forgery Localization
Image forgery localization aims to identify forged regions by capturing subtle traces from high-quality discriminative features. In this paper, we propose a Transformer-style network with two feature extraction branches for image forgery localization, and it is named as Two-Branch Transformer (TBFormer). Firstly, two feature extraction branches are elaborately designed, taking advantage of the discriminative stacked Transformer layers, for both RGB and noise domain features. Secondly, an Attention-aware Hierarchical-feature Fusion Module (AHFM) is proposed to effectively fuse hierarchical features from two different domains. Although the two feature extraction branches have the same architecture, their features have significant differences since they are extracted from different domains. We adopt position attention to embed them into a unified feature domain for hierarchical feature investigation. Finally, a Transformer decoder is constructed for feature reconstruction to generate the predicted mask. Extensive experiments on publicly available datasets demonstrate the effectiveness of the proposed model.
- 1 May 2021
- Published by Elsevier in Computers and Electrical Engineering
- Vol. 91, 107057
Learning HDR illumination from LDR panorama images
- 1 May 2021
- Published by Elsevier in Computers and Electrical Engineering
- Vol. 91, 107057
Learning HDR illumination from LDR panorama images
- Open Access
- 18 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 183647-183655
Incremental Learning of Multi-Tasking Networks for Aesthetic Radar Map Prediction
It is difficult and challenging to evaluate the aesthetics quality of images from multiple angles. Since humans’ perception of images comes from many factors, the integrated image aesthetic quality assessment cannot be easily summarized by few attributes. A comprehensive evaluation is supposed to predict many aesthetic attributes across not only one dataset. This requires the model to have not only high accuracy, but also strong generalization ability, resulting in a better prediction on multiple models and datasets. Recent work shows that deep convolution neural network can be used to extract image features and further evaluate the total score of images, and the method of evaluation are lacking of sufficient detailed features. In this paper, we propose a multi-task convolution neural network with more incremental features. We show the results in the way of a hexagon map, which is called aesthetic radar map. This allows the network model to fit different attributes in various datasets better.
- 18 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 183647-183655
Open AccessIncremental Learning of Multi-Tasking Networks for Aesthetic Radar Map Prediction
It is difficult and challenging to evaluate the aesthetics quality of images from multiple angles. Since humans’ perception of images comes from many factors, the integrated image aesthetic quality assessment cannot be easily summarized by few attributes. A comprehensive evaluation is supposed to predict many aesthetic attributes across not only one dataset. This requires the model to have not only high accuracy, but also strong generalization ability, resulting in a better prediction on multiple models and datasets. Recent work shows that deep convolution neural network can be used to extract image features and further evaluate the total score of images, and the method of evaluation are lacking of sufficient detailed features. In this paper, we propose a multi-task convolution neural network with more incremental features. We show the results in the way of a hexagon map, which is called aesthetic radar map. This allows the network model to fit different attributes in various datasets better.
- 15 June 2013
- Published by Springer Nature
- p. 2201-2208
An Approach to Verify Correctness of the Mandatory Access Control Framework
- 15 June 2013
- Published by Springer Nature
- p. 2201-2208
An Approach to Verify Correctness of the Mandatory Access Control Framework
- 1 May 2012
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 54-57
Test effectiveness index: Integrating product metrics with process metrics
Defect measurement is an important method in the improvement of software quality. Recent approaches of defect measurement are inappropriate to small software organizations by reason of their intricacy. This paper gives a simple approach of defect measurement, which integrates the power of product metrics with process metrics, i.e., it can not only detect the defect-prone modules, but also find the problems in the software process. This approach uses the results of successive two rounds of testing to create the test effectiveness index constructively. A case study is conducted and the results indicate that the defect-prone modules can be identified and problems of testing process can be discovered by test effectiveness index.
- 1 May 2012
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 54-57
Test effectiveness index: Integrating product metrics with process metrics
Defect measurement is an important method in the improvement of software quality. Recent approaches of defect measurement are inappropriate to small software organizations by reason of their intricacy. This paper gives a simple approach of defect measurement, which integrates the power of product metrics with process metrics, i.e., it can not only detect the defect-prone modules, but also find the problems in the software process. This approach uses the results of successive two rounds of testing to create the test effectiveness index constructively. A case study is conducted and the results indicate that the defect-prone modules can be identified and problems of testing process can be discovered by test effectiveness index.
- 29 November 2011
- Published by Springer Nature
- p. 561-568
Motion Object Segmentation Using Regions Classification and Energy Model
- 29 November 2011
- Published by Springer Nature
- p. 561-568
Motion Object Segmentation Using Regions Classification and Energy Model
- 1 August 2011
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 12, 2069-2072
Triangle matching combined with singular features in fingerprints
Fingerprint matching process is identical to the generation of equivalent triangular frameworks with the minutiae at their vertices. Triangle descriptor is introduced to evaluate similarity of two triangles that integrates characteristic components of the minutiae triangle with global orientation field information. To reduce computational complexity of finding correspondent triangles, fingerprint image is separated into a series of circular areas utilizing the belief function of singular points. Seed triangles are searched in limited regions that are invariant to rotation and translation and robust to distortion of fingerprints. A set of experiments have been done on different database and the better performance has been achieved.
- 1 August 2011
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 12, 2069-2072
Triangle matching combined with singular features in fingerprints
Fingerprint matching process is identical to the generation of equivalent triangular frameworks with the minutiae at their vertices. Triangle descriptor is introduced to evaluate similarity of two triangles that integrates characteristic components of the minutiae triangle with global orientation field information. To reduce computational complexity of finding correspondent triangles, fingerprint image is separated into a series of circular areas utilizing the belief function of singular points. Seed triangles are searched in limited regions that are invariant to rotation and translation and robust to distortion of fingerprints. A set of experiments have been done on different database and the better performance has been achieved.