Xiaoxia Wan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Xiaoxia Wan has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Jinxing Liang
School of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan 430200, China
6 shared publications Junfeng Li
Wuhan University
4 shared publicationsChan Li
University of Wuhan, School of Printing and Packaging Engineering, No. 129 Luoyu Street, Hongshan District, Wuhan 430079, China
3 shared publicationsQian Cao
School of Printing and Packaging, Wuhan University, Wuhan 430079, China
3 shared publicationsQiang Liu
University of Wuhan, School of Printing and Packaging Engineering, No. 129 Luoyu Street, Hongshan District, Wuhan 430079, China
3 shared publications- Junfeng Li
School of Packaging and Printing Engineering, Henan University of Animal Husbandry and Economy, Zhengzhou, China
2 shared publications
Latest Publications
- Open Access
- 15 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 129626-129634
Residual Vector Capsule: Improving Capsule by Pose Attention
The convolutional neural network has significantly improved the accuracy of image recognition; however, it performs in a fragile manner when we apply viewpoint transformation or add noise to the image. Recent studies have proposed new neural networks named capsule. Capsule build the part-whole relationship in the entity through instantiation parameters, and they cluster instantiation parameters layer by layer through routing-by-agreement; therefore, capsule has stronger representational ability and robustness than convolutional neural networks. However, the routing-by-agreement of the capsule network is limited by the prior probability assumption, which performs in an unstable way in recognition accuracy and robustness. To remove the restriction of the prior probability assumption in the routing-by-agreement, we propose a new capsule named the residual vector capsule (RVC), which constructs the routing-by-agreement with self-attention. The experimental results show that compared with other capsule networks, RVC achieves competitive classification accuracy on MNIST, Fashion-MNIST, CIFAR-10 and SVHN, improves the viewpoint invariance of the model on SmallNorb, and significantly improves the robustness of the model against white box attacks on CIFAR-10 and SVHN.
- 15 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 129626-129634
Open AccessResidual Vector Capsule: Improving Capsule by Pose Attention
The convolutional neural network has significantly improved the accuracy of image recognition; however, it performs in a fragile manner when we apply viewpoint transformation or add noise to the image. Recent studies have proposed new neural networks named capsule. Capsule build the part-whole relationship in the entity through instantiation parameters, and they cluster instantiation parameters layer by layer through routing-by-agreement; therefore, capsule has stronger representational ability and robustness than convolutional neural networks. However, the routing-by-agreement of the capsule network is limited by the prior probability assumption, which performs in an unstable way in recognition accuracy and robustness. To remove the restriction of the prior probability assumption in the routing-by-agreement, we propose a new capsule named the residual vector capsule (RVC), which constructs the routing-by-agreement with self-attention. The experimental results show that compared with other capsule networks, RVC achieves competitive classification accuracy on MNIST, Fashion-MNIST, CIFAR-10 and SVHN, improves the viewpoint invariance of the model on SmallNorb, and significantly improves the robustness of the model against white box attacks on CIFAR-10 and SVHN.
- 1 September 2020
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 29 (05) , 053006-16
Illuminant estimation using pixels spatially close to the illuminant in the rg-chromaticity space
Color constancy algorithms can provide us with illuminant invariant descriptions for a scene, and it is often accomplished by illuminant estimation. Most statistics-based methods estimate the illuminant color from the information provided by all pixels of an image. However, this research reveals that, for most images, the color of many pixels is quite different from the illuminant, and these pixels may severely trim the performance of statistics-based methods. Based on this fact, we propose a color constancy algorithm that finds a subset of image pixels with r, g components similar to those of the illuminant through a shallow neural network, and this subset of pixels is called illuminant close pixels (ICPs). Then the illuminant color is estimated from these pixels by some statistics-based methods. The proposed method has been evaluated and investigated on two benchmark datasets. Compared to using all pixels in an image, these statistics-based methods have been efficiently improved using ICPs.
- 1 September 2020
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 29 (05) , 053006-16
Illuminant estimation using pixels spatially close to the illuminant in the rg-chromaticity space
Color constancy algorithms can provide us with illuminant invariant descriptions for a scene, and it is often accomplished by illuminant estimation. Most statistics-based methods estimate the illuminant color from the information provided by all pixels of an image. However, this research reveals that, for most images, the color of many pixels is quite different from the illuminant, and these pixels may severely trim the performance of statistics-based methods. Based on this fact, we propose a color constancy algorithm that finds a subset of image pixels with r, g components similar to those of the illuminant through a shallow neural network, and this subset of pixels is called illuminant close pixels (ICPs). Then the illuminant color is estimated from these pixels by some statistics-based methods. The proposed method has been evaluated and investigated on two benchmark datasets. Compared to using all pixels in an image, these statistics-based methods have been efficiently improved using ICPs.
- 30 March 2019
- Published by Springer Nature
- p. 31-36
Study on the Constancy of Color Vision
- 2 February 2018
- Published by Springer Nature in Cellulose
- Vol. 25 (3) , 2091-2102
A modified single-constant Kubelka–Munk model for color prediction of pre-colored fiber blends
- 2 February 2018
- Published by Springer Nature in Cellulose
- Vol. 25 (3) , 2091-2102
A modified single-constant Kubelka–Munk model for color prediction of pre-colored fiber blends
- 31 October 2017
- Published by Optica Publishing Group in Optics Express
- Vol. 25 (23) , 28273-28287
Optimized method for spectral reflectance reconstruction from camera responses
An optimized method for spectral reflectance reconstruction from digital camera responses is proposed by selecting and weighting the training samples. The proposed method was based on camera responses expansion and pseudo inverse operation. Local optimal training samples were selected for each testing sample. Weighting coefficient matrix was calculated for local optimal training samples to improve the reconstruction accuracy. Experimental results indicate that the proposed method significantly outperforms currently existing methods in terms of both spectral and colorimetric accuracy, and the proposed optimized strategy was integrated well with those counterpart methods to improve their reconstruction accuracy.
- 31 October 2017
- Published by Optica Publishing Group in Optics Express
- Vol. 25 (23) , 28273-28287
Optimized method for spectral reflectance reconstruction from camera responses
An optimized method for spectral reflectance reconstruction from digital camera responses is proposed by selecting and weighting the training samples. The proposed method was based on camera responses expansion and pseudo inverse operation. Local optimal training samples were selected for each testing sample. Weighting coefficient matrix was calculated for local optimal training samples to improve the reconstruction accuracy. Experimental results indicate that the proposed method significantly outperforms currently existing methods in terms of both spectral and colorimetric accuracy, and the proposed optimized strategy was integrated well with those counterpart methods to improve their reconstruction accuracy.
- 1 March 2017
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 26 (2) , 023013
Prototype of a pigments color chart for the digital conservation of ancient murals
Digital imaging has become a very important technique in the conservation of cultural art relics because it can nondestructively acquire the color and spectral image of cultural art relics for different applications. Imaging accuracy is one of the key factors in digital protection of cultural art relics. In order to improve the color and spectral accuracy for digital imaging of cultural art relics, the idea of making the specific color charts for different kinds of artworks is presented. Taking ancient Chinese Dunhuang murals as the specific object of study, a prototype pigments color chart of the Dunhuang murals (DCC), containing a six-step grayscale and 30 colored pigment samples, is made to investigate its pigment types and painting techniques. Under the premise of considering the difference in the number of samples in color charts, the DCC is tested and compared with the classic and widely used standard Macbeth colorchecker (CC) in two aspects: color correction for RGB imaging and spectral reconstruction for spectral imaging. The results show that the prototype pigments color chart is more effective and exhibits superior performance to the CC in both aspects for digital conservation of the Dunhuang murals.
- 1 March 2017
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 26 (2) , 023013
Prototype of a pigments color chart for the digital conservation of ancient murals
Digital imaging has become a very important technique in the conservation of cultural art relics because it can nondestructively acquire the color and spectral image of cultural art relics for different applications. Imaging accuracy is one of the key factors in digital protection of cultural art relics. In order to improve the color and spectral accuracy for digital imaging of cultural art relics, the idea of making the specific color charts for different kinds of artworks is presented. Taking ancient Chinese Dunhuang murals as the specific object of study, a prototype pigments color chart of the Dunhuang murals (DCC), containing a six-step grayscale and 30 colored pigment samples, is made to investigate its pigment types and painting techniques. Under the premise of considering the difference in the number of samples in color charts, the DCC is tested and compared with the classic and widely used standard Macbeth colorchecker (CC) in two aspects: color correction for RGB imaging and spectral reconstruction for spectral imaging. The results show that the prototype pigments color chart is more effective and exhibits superior performance to the CC in both aspects for digital conservation of the Dunhuang murals.
- 18 October 2016
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 26 (1) , 011002
Nondestructive pigment size detection method of mineral paint film based on image texture
The existing methods—such as sieving, microscope, light scattering, sedimentation, and electrical induction for pigment size detection—require sampling or scattering the mineral pigments, which will inevitably cause damage to the films painted by mineral pigments. A new detection method based on run length texture analysis is proposed to nondestructively detect the pigment size in the mineral paint film. The films painted by mineral pigments with preknown pigment sizes are contactlessly captured by CCD microscope under diffused light. Gray transform, histogram equalization, and median filtering are implemented to preprocess the captured images, and then the run length texture parameters are extracted from the preprocessed images. A parametric relationship between the extracted parameters and the preknown size is established to predict the pigment size in mineral paint film nondestructively. Burnt carnelian is selected as the sample to verify the feasibility of the proposed method. Results show that the max detection error of the proposed method is and can be applied to the size detection of the mineral pigments used in mineral paint film.- 18 October 2016
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 26 (1) , 011002
Nondestructive pigment size detection method of mineral paint film based on image texture
The existing methods—such as sieving, microscope, light scattering, sedimentation, and electrical induction for pigment size detection—require sampling or scattering the mineral pigments, which will inevitably cause damage to the films painted by mineral pigments. A new detection method based on run length texture analysis is proposed to nondestructively detect the pigment size in the mineral paint film. The films painted by mineral pigments with preknown pigment sizes are contactlessly captured by CCD microscope under diffused light. Gray transform, histogram equalization, and median filtering are implemented to preprocess the captured images, and then the run length texture parameters are extracted from the preprocessed images. A parametric relationship between the extracted parameters and the preknown size is established to predict the pigment size in mineral paint film nondestructively. Burnt carnelian is selected as the sample to verify the feasibility of the proposed method. Results show that the max detection error of the proposed method is and can be applied to the size detection of the mineral pigments used in mineral paint film.- 26 August 2016
- Published by Optica Publishing Group in Journal of the Optical Society of America A
- Vol. 33 (9) , 1860-1871
Updated version of an interim connection space LabPQR for spectral color reproduction: LabLab
In this paper, we propose a new interim connection space (ICS) called LabLab, which is an updated version of LabPQR, to overcome the drawback that the last three dimensions of LabPQR have no definite colorimetric meanings. We extended and improved the method by which the first three dimensions of LabPQR are deduced to obtain an ICS consisting of two sets of CIELAB values under different illuminants, and the reconstructed spectra from LabLab were obtained by minimizing colorimetric errors by means of the computational formula of the CIE-XYZ tristimulus values combined with least-squares best fit. The improvement obtained from the proposed method was tested to compress and reconstruct the reflectance spectra of the 1950 Natural Color System color chips and more than 50,000 ISO SOCS color patches as well as six multispectral images acquired by multispectral image acquisition systems using 1600 glossy Munsell color chips as training samples. The performance was evaluated by the mean values of color differences between the original and reconstructed spectra under the CIE 1931 standard colorimetric observer and the CIE standard illuminants D50, D55, D65, D75, F2, F7, F11, and A as well as five multichip white LED light sources. The mean and maximum values of the root mean square errors between the original and reconstructed spectra were also calculated. The experimental results show that the proposed three LabLab interim connection spaces significantly outperform principal component analysis, LabPQR, XYZLMS, Fairman–Brill, and LabRGB in colorimetric reconstruction accuracy at the cost of slight reduction of spectral reconstruction accuracy and illuminant independence of color differences of the suggested LabLab interim connection spaces outperform other interim connection spaces. In addition, the presented LabLab interim connection spaces could be quite compatible with the extensively used colorimetric management system since each dimension has definite colorimetric meanings and is perceptually uniform.
- 26 August 2016
- Published by Optica Publishing Group in Journal of the Optical Society of America A
- Vol. 33 (9) , 1860-1871
Updated version of an interim connection space LabPQR for spectral color reproduction: LabLab
In this paper, we propose a new interim connection space (ICS) called LabLab, which is an updated version of LabPQR, to overcome the drawback that the last three dimensions of LabPQR have no definite colorimetric meanings. We extended and improved the method by which the first three dimensions of LabPQR are deduced to obtain an ICS consisting of two sets of CIELAB values under different illuminants, and the reconstructed spectra from LabLab were obtained by minimizing colorimetric errors by means of the computational formula of the CIE-XYZ tristimulus values combined with least-squares best fit. The improvement obtained from the proposed method was tested to compress and reconstruct the reflectance spectra of the 1950 Natural Color System color chips and more than 50,000 ISO SOCS color patches as well as six multispectral images acquired by multispectral image acquisition systems using 1600 glossy Munsell color chips as training samples. The performance was evaluated by the mean values of color differences between the original and reconstructed spectra under the CIE 1931 standard colorimetric observer and the CIE standard illuminants D50, D55, D65, D75, F2, F7, F11, and A as well as five multichip white LED light sources. The mean and maximum values of the root mean square errors between the original and reconstructed spectra were also calculated. The experimental results show that the proposed three LabLab interim connection spaces significantly outperform principal component analysis, LabPQR, XYZLMS, Fairman–Brill, and LabRGB in colorimetric reconstruction accuracy at the cost of slight reduction of spectral reconstruction accuracy and illuminant independence of color differences of the suggested LabLab interim connection spaces outperform other interim connection spaces. In addition, the presented LabLab interim connection spaces could be quite compatible with the extensively used colorimetric management system since each dimension has definite colorimetric meanings and is perceptually uniform.
- 12 August 2016
- Published by Springer Nature in Optical Review
- Vol. 23 (5) , 753-764
Spectral data compression using weighted principal component analysis with consideration of human visual system and light sources
- 12 August 2016
- Published by Springer Nature in Optical Review
- Vol. 23 (5) , 753-764
Spectral data compression using weighted principal component analysis with consideration of human visual system and light sources
- 28 July 2016
- Published by SAGE Publications in Applied Spectroscopy
- Vol. 70 (11) , 1900-1909
In Situ Identification of Pigment Composition and Particle Size on Wall Paintings Using Visible Spectroscopy as a Noninvasive Measurement Method
Noninvasive examination methods of chemical composition and particle size are presented here based on visible spectroscopy to achieve the identification and recording of mineral pigments used on ancient wall paintings. The normalized spectral curve, slope and curvature extracted from visible spectral reflectance are combined with adjustable weighting coefficients to construct the identification feature space, and Euclid distances between spectral reflectance from wall paintings and a reference database are calculated in the feature space as the discriminant criterion to identify the chemical composition of mineral pigments. A parametric relationship between the integral quantity of spectral reflectance and logarithm of mean particle size is established using a quadratic polynomial to accomplish the noninvasive prediction of mineral pigment particle size used on ancient wall paintings. The feasibility of the proposed methods is validated by the in situ nondestructive identification of the wall paintings in the Mogao Grottoes at Dunhuang. Chinese painting styles and historical evolution are then analyzed according to the identification results of 16 different grottoes constructed from the Sixteen Kingdoms to the Yuan Dynasty.
- 28 July 2016
- Published by SAGE Publications in Applied Spectroscopy
- Vol. 70 (11) , 1900-1909
In Situ Identification of Pigment Composition and Particle Size on Wall Paintings Using Visible Spectroscopy as a Noninvasive Measurement Method
Noninvasive examination methods of chemical composition and particle size are presented here based on visible spectroscopy to achieve the identification and recording of mineral pigments used on ancient wall paintings. The normalized spectral curve, slope and curvature extracted from visible spectral reflectance are combined with adjustable weighting coefficients to construct the identification feature space, and Euclid distances between spectral reflectance from wall paintings and a reference database are calculated in the feature space as the discriminant criterion to identify the chemical composition of mineral pigments. A parametric relationship between the integral quantity of spectral reflectance and logarithm of mean particle size is established using a quadratic polynomial to accomplish the noninvasive prediction of mineral pigment particle size used on ancient wall paintings. The feasibility of the proposed methods is validated by the in situ nondestructive identification of the wall paintings in the Mogao Grottoes at Dunhuang. Chinese painting styles and historical evolution are then analyzed according to the identification results of 16 different grottoes constructed from the Sixteen Kingdoms to the Yuan Dynasty.