Xi Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Xi Wu has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Zhenwei Shi
Beihang University
2 shared publications Asvin Ananthanarayan
Microsoft
1 shared publicationAtul Shenoy
Microsoft
1 shared publication- Edward Gan
Stanford InfoLab, Stanford, CA, USA
1 shared publication Eric Xu
Microsoft
1 shared publication- Erik Meijer
Center for Economic and Social Research, University of Southern California , Los Angeles, California
1 shared publication
Latest Publications
- 1 February 2022
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 36 (02)
Person Re-Identification Combined with Style Transfer and Pose Generation
The number of existing person re-identification datasets is limited, and there are a series of changes such as illumination, background occlusion and pose among each dataset, which makes it difficult for the existing methods to learn robust feature representation, leading to a decline in recognition performance. To solve these problems, a person re-identification method combining style and pose generation is proposed in this paper. First with the impact of camera style differences in collecting images from different cameras, a style transformation method based on generative adversarial network is introduced into a person re-identification model, and cyclic generative adversarial networks (CycleGAN) is used to realize style transfer and reduce the influence of camera differences. Second, in view of the problem that when pedestrian pose changes greatly, easy to ignore identity-sensitive related information, AlphaPose is introduced to implement pose estimation. Combining style and posture for the first time and using improved deep convolution generative adversarial network (DCGAN) structure enrich the input sample information and generate unified style pose image; while using new synthetic data to train person re-identification network model improves the recognition performance of the model. Finally, further introducing random erasure method during data enhancement, in order to reduce the overfitting phenomena, improves the generalization ability of the network simultaneously and solves partial occlusion. The experimental results show that the proposed method outperforms typical style-based or pose-based methods. The accuracy of rank-1 and mAP on Market-1501 dataset is 90.4% and 74.5%, respectively, which are 2.28% and 5.78% higher, respectively. To a certain extent, the performance of person re-identification is improved.
- 1 February 2022
- Published by World Scientific Pub Co Pte Ltd in International Journal of Pattern Recognition and Artificial Intelligence
- Vol. 36 (02)
Person Re-Identification Combined with Style Transfer and Pose Generation
The number of existing person re-identification datasets is limited, and there are a series of changes such as illumination, background occlusion and pose among each dataset, which makes it difficult for the existing methods to learn robust feature representation, leading to a decline in recognition performance. To solve these problems, a person re-identification method combining style and pose generation is proposed in this paper. First with the impact of camera style differences in collecting images from different cameras, a style transformation method based on generative adversarial network is introduced into a person re-identification model, and cyclic generative adversarial networks (CycleGAN) is used to realize style transfer and reduce the influence of camera differences. Second, in view of the problem that when pedestrian pose changes greatly, easy to ignore identity-sensitive related information, AlphaPose is introduced to implement pose estimation. Combining style and posture for the first time and using improved deep convolution generative adversarial network (DCGAN) structure enrich the input sample information and generate unified style pose image; while using new synthetic data to train person re-identification network model improves the recognition performance of the model. Finally, further introducing random erasure method during data enhancement, in order to reduce the overfitting phenomena, improves the generalization ability of the network simultaneously and solves partial occlusion. The experimental results show that the proposed method outperforms typical style-based or pose-based methods. The accuracy of rank-1 and mAP on Market-1501 dataset is 90.4% and 74.5%, respectively, which are 2.28% and 5.78% higher, respectively. To a certain extent, the performance of person re-identification is improved.
- 26 October 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
Scene Aggregation Network for Cloud Detection on Remote Sensing Imagery
There has been a breakthrough in cloud detection by using convolutional neural networks (CNNs) during these years. However, there are still weaknesses among current cloud detection algorithms because only cloud mask information is used. As clouds represent differently in different scenes, the scene information may give hints to improve cloud detection performance. Therefore, different from the previous cloud detection literature, in this letter, we propose an end-to-end new deep learning network named scene aggregation network (SAN), which aggregates the scene information in the framework. Specifically, basic features are first extracted by utilizing all levels of network features. Then, the aggregated features used to produce the final cloud masks are created by fusing the basic features and the specially introduced scene information. Experimental results have demonstrated that with scene information aggregated, our proposed method can be robust on images with different scenes. Additionally, as SAN outperforms other state-of-the-art methods, our proposed method suits for cloud detection and can achieve improvement on this task.
- 26 October 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
Scene Aggregation Network for Cloud Detection on Remote Sensing Imagery
There has been a breakthrough in cloud detection by using convolutional neural networks (CNNs) during these years. However, there are still weaknesses among current cloud detection algorithms because only cloud mask information is used. As clouds represent differently in different scenes, the scene information may give hints to improve cloud detection performance. Therefore, different from the previous cloud detection literature, in this letter, we propose an end-to-end new deep learning network named scene aggregation network (SAN), which aggregates the scene information in the framework. Specifically, basic features are first extracted by utilizing all levels of network features. Then, the aggregated features used to produce the final cloud masks are created by fusing the basic features and the specially introduced scene information. Experimental results have demonstrated that with scene information aggregated, our proposed method can be robust on images with different scenes. Additionally, as SAN outperforms other state-of-the-art methods, our proposed method suits for cloud detection and can achieve improvement on this task.
- Open Access
- 1 December 2018
- Published by Association for Computing Machinery (ACM) in Proceedings of the VLDB Endowment
- Vol. 12 (4) , 419-432
DIFF
A range of explanation engines assist data analysts by performing feature selection over increasingly high-volume and high-dimensional data, grouping and highlighting commonalities among data points. While useful in diverse tasks such as user behavior analytics, operational event processing, and root cause analysis, today's explanation engines are designed as standalone data processing tools that do not interoperate with traditional, SQL-based analytics workflows; this limits the applicability and extensibility of these engines. In response, we propose the DIFF operator, a relational aggregation operator that unifies the core functionality of these engines with declarative relational query processing. We implement both single-node and distributed versions of the DIFF operator in MB SQL, an extension of MacroBase, and demonstrate how DIFF can provide the same semantics as existing explanation engines while capturing a broad set of production use cases in industry, including at Microsoft and Facebook. Additionally, we illustrate how this declarative approach to data explanation enables new logical and physical query optimizations. We evaluate these optimizations on several real-world production applications, and find that DIFF in MB SQL can outperform state-of-the-art engines by up to an order of magnitude.
- 1 December 2018
- Published by Association for Computing Machinery (ACM) in Proceedings of the VLDB Endowment
- Vol. 12 (4) , 419-432
Open AccessDIFF
A range of explanation engines assist data analysts by performing feature selection over increasingly high-volume and high-dimensional data, grouping and highlighting commonalities among data points. While useful in diverse tasks such as user behavior analytics, operational event processing, and root cause analysis, today's explanation engines are designed as standalone data processing tools that do not interoperate with traditional, SQL-based analytics workflows; this limits the applicability and extensibility of these engines. In response, we propose the DIFF operator, a relational aggregation operator that unifies the core functionality of these engines with declarative relational query processing. We implement both single-node and distributed versions of the DIFF operator in MB SQL, an extension of MacroBase, and demonstrate how DIFF can provide the same semantics as existing explanation engines while capturing a broad set of production use cases in industry, including at Microsoft and Facebook. Additionally, we illustrate how this declarative approach to data explanation enables new logical and physical query optimizations. We evaluate these optimizations on several real-world production applications, and find that DIFF in MB SQL can outperform state-of-the-art engines by up to an order of magnitude.
- Open Access
- 21 November 2018
- Published by MDPI AG in Remote Sensing
- Vol. 10 (11) , 1853
Utilizing Multilevel Features for Cloud Detection on Satellite Imagery
Cloud detection, which is defined as the pixel-wise binary classification, is significant in satellite imagery processing. In current remote sensing literature, cloud detection methods are linked to the relationships of imagery bands or based on simple image feature analysis. These methods, which only focus on low-level features, are not robust enough on the images with difficult land covers, for clouds share similar image features such as color and texture with the land covers. To solve the problem, in this paper, we propose a novel deep learning method for cloud detection on satellite imagery by utilizing multilevel image features with two major processes. The first process is to obtain the cloud probability map from the designed deep convolutional neural network, which concatenates deep neural network features from low-level to high-level. The second part of the method is to get refined cloud masks through a composite image filter technique, where the specific filter captures multilevel features of cloud structures and the surroundings of the input imagery. In the experiments, the proposed method achieves 85.38% intersection over union of cloud in the testing set which contains 100 Gaofen-1 wide field of view images and obtains satisfactory visual cloud masks, especially for those hard images. The experimental results show that utilizing multilevel features by the combination of the network with feature concatenation and the particular filter tackles the cloud detection problem with improved cloud masks.
- 21 November 2018
- Published by MDPI AG in Remote Sensing
- Vol. 10 (11) , 1853
Open AccessUtilizing Multilevel Features for Cloud Detection on Satellite Imagery
Cloud detection, which is defined as the pixel-wise binary classification, is significant in satellite imagery processing. In current remote sensing literature, cloud detection methods are linked to the relationships of imagery bands or based on simple image feature analysis. These methods, which only focus on low-level features, are not robust enough on the images with difficult land covers, for clouds share similar image features such as color and texture with the land covers. To solve the problem, in this paper, we propose a novel deep learning method for cloud detection on satellite imagery by utilizing multilevel image features with two major processes. The first process is to obtain the cloud probability map from the designed deep convolutional neural network, which concatenates deep neural network features from low-level to high-level. The second part of the method is to get refined cloud masks through a composite image filter technique, where the specific filter captures multilevel features of cloud structures and the surroundings of the input imagery. In the experiments, the proposed method achieves 85.38% intersection over union of cloud in the testing set which contains 100 Gaofen-1 wide field of view images and obtains satisfactory visual cloud masks, especially for those hard images. The experimental results show that utilizing multilevel features by the combination of the network with feature concatenation and the particular filter tackles the cloud detection problem with improved cloud masks.