Shilin Zhou

Shilin Zhou

Auto-generated profile, based on publicly available information

Shilin Zhou

College of Electronic Science and Technology, National University of Defense Technology, China

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

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

Representative Publications

View All Publications
6 May 2017 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Closed access
ZDZhipeng Deng
Zhipeng Deng
HSHao Sun
Hao Sun
SZShilin Zhou
Shilin Zhou
JZJuanping Zhao
Juanping Zhao
HZHuanxin Zou
Huanxin Zou
Vehicle detection in aerial images, being an interesting but challenging problem, plays an important role for a wide range of applications. Traditional methods are based on sliding-window search and handcrafted or shallow-learning-based features with heavy computational costs and limited representation power. Recently, deep learning algorithms, especially region-based convolutional neural networks (R-CNNs), have achieved state-of-the-art detection performance in computer vision. However, several challenges limit the applications of R-CNNs in vehicle detection from aerial images: 1) vehicles in large-scale aerial images are relatively small in size, and R-CNNs have poor localization performance with small objects; 2) R-CNNs are particularly designed for detecting the bounding box of the targets without extracting attributes; 3) manual annotation is generally expensive and the available manual annotation of vehicles for training R-CNNs are not sufficient in number. To address these problems, this paper proposes a fast and accurate vehicle detection framework. On one hand, to accurately extract vehicle-like targets, we developed an accurate-vehicle-proposal-network (AVPN) based on hyper feature map which combines hierarchical feature maps that are more accurate for small object detection. On the other hand, we propose a coupled R-CNN method, which combines an AVPN and a vehicle attribute learning network to extract the vehicle's location and attributes simultaneously. For original large-scale aerial images with limited manual annotations, we use cropped image blocks for training with data augmentation to avoid overfitting. Comprehensive evaluations on the public Munich vehicle dataset and the collected vehicle dataset demonstrate the accuracy and effectiveness of the proposed method.

1 June 2018
Verified
Conference Paper
Closed access
XZXiaoli Zhao
Xiaoli Zhao
SZShilin Zhou
Shilin Zhou
LLLin Lei
Lin Lei
ZDZhipeng Deng
Zhipeng Deng
In Unmanned Aerial Vehicle (UAV) videos, object tracking remains a challenge, due to its low spatial resolution and poor real-time performance. Recently, methods of deep learning have made great progress in object tracking in computer vision, especially fully-convolutional siamese neural networks (SiamFC). Inspired by it, this paper aims to investigate the use of SiamFC for object tracking in UAV videos. The network is trained on part of a UAV123 dataset and Stanford Drone dataset. First, exemplar image is extracted from the first frame and search regions are extracted in the following frames. Then, a Siamese network is used for tracking objects by calculating the similarity between exemplar image and search region. To evaluate our method, we test on a challenge VIVID dataset. The experiment shows that the proposed method has improvements in accuracy and speed in low spatial resolution UAV videos compared to existing methods.

1 July 2012 in Electrochimica Acta
Verified
Journal Article
Closed access
HWHuihui Wang
Huihui Wang
SZShilin Zhou
Shilin Zhou
ZWZhigang Wang
Zhigang Wang
SXShanling Xu
Shanling Xu
SDSong Dong
Song Dong
YMYuqing Miao
Yuqing Miao

1 January 2013 in International Journal of Advanced Robotic Systems
Verified
Journal Article
Open access
HSHao Sun
Hao Sun
ZHZou Huanxin
Zou Huanxin
SZShilin Zhou
Shilin Zhou
CWCheng Wang
Cheng Wang
Naser El-SheimyNaser El-Sheimy
Naser El-Sheimy
Detection and tracking surrounding moving obstacles such as vehicles and pedestrians are crucial for the safety of mobile robotics and autonomous vehicles. This is especially the case in urban driving scenarios. This paper presents a novel framework for surrounding moving obstacles detection using binocular stereo vision. The contributions of our work are threefold. Firstly, a multiview feature matching scheme is presented for simultaneous stereo correspondence and motion correspondence searching. Secondly, the multiview geometry constraint derived from the relative camera positions in pairs of consecutive stereo views is exploited for surrounding moving obstacles detection. Thirdly, an adaptive particle filter is proposed for tracking of multiple moving obstacles in surrounding areas. Experimental results from real-world driving sequences demonstrate the effectiveness and robustness of the proposed framework.

1 November 2021
Verified
Conference Paper
Closed access
ZFZuojun Fu
Zuojun Fu
YHYi Hou
Yi Hou
CLCangjian Liu
Cangjian Liu
YZYi Zhang
Yi Zhang
SZShilin Zhou
Shilin Zhou
In recent years, autonomous driving solutions that rely on the fusion of multiple sensors such as lidar, inertial navigation systems, and GPS positioning systems have attracted the attention of the industry and academia. Such solutions use multiple sets of sensing systems to work together to ensure the greatest probability of the accuracy and completeness of the results. These solutions are able to tackle problems that are difficult to solve with a single sensor in the short term. However, in the long run, the design of deep coupling between data and strategy is not conducive to the real improvement of the perception system. Therefore, this paper proposes a light-weight autonomous vehicle system based on pure vision named VOLWA. To cope with the huge challenge of using the camera as the only sensor, VOLWA contains three new algorithms:(i) view segmentation algorithm is proposed to improve the perception accuracy (ii) visual place recognition algorithm for removing dynamic targets based on channel selection is proposed to solve the interference problem of dynamic targets such as pedestrians and vehicles in the field of view (iii) threshold self-localization based on image features is proposed for high-precision positioning of autonomous vehicles. Experimental results show that visual place recognition algorithm for removing dynamic targets based on channel selection effectively improves the performance of place recognition. The system architecture and algorithm proposed in this paper have been transplanted to self-developed autonomous vehicles for verification, and the autonomous vehicles are running well. See the link for the demo video https://youtu.be/6WfUnor205M.

19 May 2023
Verified
Conference Paper
Open access
LQLiangcheng Qin
Liangcheng Qin
ZSZhichao Sha
Zhichao Sha
SZShilin Zhou
Shilin Zhou
WDWei Du
Wei Du
YGYulan Guo
Yulan Guo
YZYaoyuan Zeng
Yaoyuan Zeng
Feature aggregation has an important impact on semantic segmentation, especially the segmentation of object boundaries. With recent advances of various feature aggregation techniques, great progress has been achieved in semantic segmentation. However, it is still a challenge to effectively incorporate high-level semantic features and low-level spatial features. In this work, we propose the Class-Guide Deformable Alignment Fusion network (CGDAFNet) for real-time semantic segmentation. Instead of simply concatenating or adding cross-level features directly, we propose a class-guide deformable alignment fusion module (CGDAFM) to handle the misalignment problem between cross-level features, so as to achieve effective cross-level feature incorporation. A grouped double normalization residual module (GDNRM) is proposed to improve segmentation performance with a small computation cost. Experiment results on the CamVid and Cityscapes datasets verify the effectiveness of our method. Our CGDAFNet can achieve a mIoU score of 78.8% on Cityscapes dataset by taking input with the resolution of 1024×2048. In particular, the speed of our method is 39 FPS with single RTX 3080.