Yule Zhang

Yule Zhang

Auto-generated profile, based on publicly available information

Yule Zhang

Graduate College, Air Force Engineering University, Xi'an, China

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 May 2021 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
ZMZhu mingming Zhu mingming
Zhu mingming Zhu mingming
GHGuoping Hu
Guoping Hu
HZHao Zhou
Hao Zhou
SWShiqiang Wang
Shiqiang Wang
YZYule Zhang
Yule Zhang
KZKexin Zang
Kexin Zang
Aiming to address the problems of arbitrary orientations, large aspect ratios, and dense arrangements in ship detection, an arbitrary-oriented ship detection method based on RetinaNet is proposed. Our proposed method includes a rotated RetinaNet, a refined network, a feature alignment module, and an improved loss function. First, the rotated RetinaNet achieves rotation detection by using a feature pyramid network, rotated anchors, the skew intersection-over-union (IoU), and skew nonmaximum suppression. Then, the refined network and feature alignment module are introduced to achieve better detection accuracy. Finally, to address the boundary discontinuity, the loss function is improved by introducing the IoU constant factor. Considering the problems with the HRSC2016 dataset, we establish a new dataset with more accurate labels and more images and object samples. Through an ablation study, we thoroughly analyze the validity of the proposed rotated RetinaNet, feature alignment module, and improved loss function. The experimental results show that our method is superior to other state-of-the-art methods.

1 March 2021 in Science of The Total Environment
Verified
Journal Article
Closed access
LZLulu Zheng
Lulu Zheng
YWYuwen Wang
Yuwen Wang
YZYule Zhang
Yule Zhang
YFYongfeng Fu
Yongfeng Fu
ZYZhijin Yang
Zhijin Yang
SLSixiu Liu
Sixiu Liu

10 February 2021 in IEEE Access
Verified
Journal Article
Open access
ZMZhu mingming Zhu mingming
Zhu mingming Zhu mingming
GHGuoping Hu
Guoping Hu
SLShuai Li
Shuai Li
HZHao Zhou
Hao Zhou
SWShiqiang Wang
Shiqiang Wang
SYShijie Yue
Shijie Yue
To address these problems, namely, arbitrary orientations, various sizes, and dense distributions of ship detection, we propose an arbitrary-oriented ship detection method via rotated single-stage detector (ROS-Det), which integrates a feature pyramid network (FPN) based on an improved ResNet50, rotated anchors, a classification network, and a regression network together. Firstly, to improve robustness against various sizes of ships, the FPN is used to fusion multiscale convolutional feature maps. Through several tweaks, the improved ResNet50 can receive more information and reduce the computational cost. Secondly, for the purpose of arbitrary-oriented ship detection, rotated anchors, skew intersection over union (IoU), and skew non-maximum suppression (NMS) are introduced to RetinaNet. Then, on account of the disadvantages that the arbitrary-oriented object detection methods usually cause loss discontinuity problem, we improve the traditional smooth L1 loss function by introducing an IoU constant factor. Finally, based on several techniques such as data augmentation and transfer learning, we achieve ship detection on a public ship dataset HRSC2016. Through comparison experiments, we have analyzed and discussed the validity of our proposed ROS-Det, which achieves the state-of-the-art performance.

1 September 2017 in BJR|case reports
Verified
Journal Article
Open access
YZYule Zhang
Yule Zhang
GHGuohua Huang
Guohua Huang
WXWeiqiang Xiao
Weiqiang Xiao
LZLiling Zhu
Liling Zhu
NWNa Wang
Na Wang
BGBuyun Guan
Buyun Guan
Quantitative dynamic contrast enhanced MRI (DCE-MRI) can offer information related to tumour perfusion and permeability (Ktrans), rate constant (Kep), extravascular extracellular volume fraction (Ve) and distribution volume (Vd). Different types of gadolinium-based contrast agents (GBCAs) may traverse the vascular wall with different velocities owing to their physicochemical characteristics. The purpose of this article was to compare the DCE-MRI quantitative results (Ktrans, Kep, Ve and Vd) between Magnevist and Eovist in a VX2 rabbit liver tumour model. Sixteen rabbits (body weight, 3 Kg; random gender) containing implanted hepatic VX2 carcinomas were randomly divided into two groups based on the regimen of MRI contrast agent administered, eight rabbits in each group. All rabbits underwent a liver DCE-MRscan before tumour transplantation. Fourteen days after tumour transplantation, the eight rabbits in Group A (Magnevist group) underwent a liver DCE-MR scan in a 3.0 T Magnetom Verio MR scanner (Siemens Healthcare, AD, Germany) after the administration of Magnevist at the flow rate of 1 ml s–1. The Group B rabbits underwent the same scan except for the administration of Eovist at the same flow rate. Twenty-four hours after the initial DCE-MRI, repeat DCE-MRI was performed with the cross-over GBCA at the same flow rate in each group. Every rabbit received 0.6 ml GBCA (0.2 ml Kg–1) during each DCE-MRI. Ktrans, Kep, Ve and Vd were measured in the tumour lesion and compared with normal liver tissue in the same slice. A pathologic examination was also performed. Hepatocellular carcinoma was diagnosed in all 16 rabbits by pathologic examination. There were no significant differences in Ktrans, Ve, Kep and Vd between the two groups of rabbits (p > 0.05). The Ktrans, Ve, Kep and Vd of the VX2 rabbit liver tumour model were significantly higher than the normal liver parenchyma (0.742 ± 0.086 vs 0.027 ± 0.002, 7.345 ± 0.043 vs 6.721 ± 0.035, 0.101 ± 0.005 vs 0.101 ± 0.005, 0.419 ± 0.083 vs 0.037 ± 0.005, respectively; p < 0.01). The Ktrans, Ve and Vd of Eovist group were significantly higher compared with the values in the Magnevist group (0.116 ± 0.016 vs 0.010 ± 0.002, respectively, p < 0.01; 0.101 ± 0.005 vs 0.004 ± 0.0009, respectively, p < 0.01; 0.419 ± 0.083 vs 0.037 ± 0.005, respectively, p < 0.001). There was no significant difference in Kep between the Eovist and Magnevist groups (7.345 ± 0.043 vs 6.721 ± 0.035, respectively; p > 0.05). In the VX2 rabbit liver tumour model, DCE-MRI performed with different types of GBCA can develop different quantitative results with respect to Ktrans, Ve and Vd. The liver-specific GBCA, Eovist, is more sensitive than the general GBCA, Magnevist, in detecting tumour perfusion and permeability.