Proceedings of the 2nd International Conference on Graphics and Signal Processing最新文献_第2页

Pulmonary Nodule Detection in CT Images via Deep Neural Network: Nodule Candidate Detection 基于深度神经网络的CT图像肺结节检测:结节候选检测

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282302

Zhengwei Hu, A. Muhammad, Ming Zhu

{"title":"Pulmonary Nodule Detection in CT Images via Deep Neural Network: Nodule Candidate Detection","authors":"Zhengwei Hu, A. Muhammad, Ming Zhu","doi":"10.1145/3282286.3282302","DOIUrl":"https://doi.org/10.1145/3282286.3282302","url":null,"abstract":"In recent years nodule candidate detection becomes the basis of the automated pulmonary nodule detection system, of which the the upper bound limit performance is determined by the sensitivity of nodule candidates detection. This paper is to improve the nodule candidate detection using deep neural networks. We treat the nodule detection task as pixel-level segmentation problem. Based on the 2D U-NET network. We build a multi-level network to process each CT slice to detect more nodules. Weighted dice loss function is designed to maintain a high sensitivity. More important, different from normaly segmentation problem, it has a heavily unbalanced positive and negative samples. We proposed a training method to make the network converge easily. We further propose an effective non-maximum suppression (NMS) method to remove duplicate nodules. The proposed framework has been validated on LUNA16 dataset. We achieved 94.3% sensitivity score, and had a 1/3 times of false positives less than the official methods of LUNA which is better for false positive reduction task. We provide a deep neural network solution for nodule candidate detection and the experimental result demonstrates the effectiveness of our method. It can also be used for input of the false positive reduction task.","PeriodicalId":324982,"journal":{"name":"Proceedings of the 2nd International Conference on Graphics and Signal Processing","volume":"139 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127475638","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 9

An Improved Method for Person Re-identification 一种改进的人员再识别方法

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282301

Han Jiang, Xinmei Yang, Yaobin Li

{"title":"An Improved Method for Person Re-identification","authors":"Han Jiang, Xinmei Yang, Yaobin Li","doi":"10.1145/3282286.3282301","DOIUrl":"https://doi.org/10.1145/3282286.3282301","url":null,"abstract":"This paper proposes a new method which combine Singular Vector Decomposition with k-reciprocal encoding for the application of person re-identification(re-ID). When we use the Euclidean distance to retrieve person, it is observed that the weight vectors in a fully connected layer are usually correlated, which makes a large impact on the retrieval result. Singular Vector Decomposition is adopted to decorrelation in this article, which has a better performance with the restraint and relaxation iteration training. Meanwhile, we add a k-reciprocal method to above result, our hypothesis is based on a gallery image is more likely to match the probe when they are in the k-reciprocal nearest neighbors. So we combine a k-reciprocal feature which is calculated by encoding its k-reciprocal nearest neighbors into a single vector under Jaccard distance and original distance as the final distance. Our method has been experimented on Market-1501 and CUHK03, it achieves a great performance, the results show that, rank-1 accuracy is improved to 82.69% and mAP is improved to 70.60% on Market-1501 for CaffeNet, while for ResNet-50, rank-1 accuracy is improved to 82.63% and mAP is improved to 73.32%.","PeriodicalId":324982,"journal":{"name":"Proceedings of the 2nd International Conference on Graphics and Signal Processing","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115904031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Database Construction and SPC Technology for Solder Paste Defect Inspection 锡膏缺陷检测数据库建设与SPC技术

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282304

Min Qi, Xin Chang, Yuelei Xu, Lechi Zhang

引用次数: 0

A Novel Binocular Stereo Vision-Based Approach of Cycle Path Detection 一种新的基于双目立体视觉的循环路径检测方法

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282295

Dezhong Tong, Geng Liu, Xiangpeng Liu

引用次数: 0

Vehicle Speed Measurement Based on Camera Calibration 基于摄像机标定的车辆速度测量

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282288

Yuqing Song, Z. Guan, Wenzhang He

{"title":"Vehicle Speed Measurement Based on Camera Calibration","authors":"Yuqing Song, Z. Guan, Wenzhang He","doi":"10.1145/3282286.3282288","DOIUrl":"https://doi.org/10.1145/3282286.3282288","url":null,"abstract":"Road traffic injuries cause considerable economic losses to individuals, their families, and to nations as a whole. Every year the lives of more than 1.25 million people are cut short as a result of a road traffic crash. A common cause of accidents is driving faster than one can stop within their field of vision. Measuring a vehicle's speed using video data is an important technique in determining the cause of a traffic accident. When on-site distance measurement cannot be carried out for a video recorded by a surveillance camera, the speed of a vehicle moving along a straight line in the video can be measured by selecting 2 reference points on the body of the vehicle and computing the speed of the vehicle during the time when the part of the vehicle between the reference points passed an environment point. The 2-reference-point method is widely used in traffic accident investigations in China. However, in real-world applications it often lacks high accuracy. In this paper we propose an improved method based on camera calibration. The proposed method selects 4 coplanar points on the body of the vehicle and computes their world coordinates by camera calibration. The speed is measured based on the movements of the 4 points. Experiments confirmed that the proposed method outperforms the 2-reference-point method.","PeriodicalId":324982,"journal":{"name":"Proceedings of the 2nd International Conference on Graphics and Signal Processing","volume":"28 22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129215011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

A Discussion on the Practice of Government New Media in the Construction of Chinese Government Websites 论政府新媒体在我国政府网站建设中的实践

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282306

Li-hai Cheng, Jun Cheng, Li-Wei Zhu, Wen Geng

引用次数: 2

Linear Camera Calibration by Single Image Based on Distortion Correction 基于畸变校正的单图像线性摄像机标定

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282303

Min Qi, B. Zhang, Yuelei Xu, Hongjuan Xin, Gong Cheng

{"title":"Linear Camera Calibration by Single Image Based on Distortion Correction","authors":"Min Qi, B. Zhang, Yuelei Xu, Hongjuan Xin, Gong Cheng","doi":"10.1145/3282286.3282303","DOIUrl":"https://doi.org/10.1145/3282286.3282303","url":null,"abstract":"In this paper, a linear camera calibration algorithm using one image based on distortion correction is presented. First, this algorithm takes advantage of first-order radial distortion coefficient to correct image primary and then extracts the control points to fit line based on orthogonal distance, getting accurate linear slope. Next, according to the slopes of collinear points, distortion correction index function is established and then all distortions parameters are obtained by optimizing the function using optimization algorithm. Finally, camera can be calibrated linearly with single image that have been corrected based on Zhengyou Zhang algorithm. Distortion parameters, separated from the camera model, do not need to be incorporated into the camera model for multiple repeated calibrations. On the one hand, it solves the problem that Zhengyou Zhang algorithm can not be used in camera systems with a low depth of field. On the other hand, it avoids the instability of solution caused by coupling of distortion parameters, internal parameters and external parameters as well as the complex computations of three LM optimization. A large number of analysis and experiments prove that this algorithm is simple, efficient and has wide application in theory and practice.","PeriodicalId":324982,"journal":{"name":"Proceedings of the 2nd International Conference on Graphics and Signal Processing","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131182972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

Deep Neural Networks Optimization Based On Deconvolutional Networks 基于反卷积网络的深度神经网络优化

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282299

Zhoufeng Liu, Chi Zhang, Chunlei Li, S. Ding, Shanliang Liu, Yan Dong

{"title":"Deep Neural Networks Optimization Based On Deconvolutional Networks","authors":"Zhoufeng Liu, Chi Zhang, Chunlei Li, S. Ding, Shanliang Liu, Yan Dong","doi":"10.1145/3282286.3282299","DOIUrl":"https://doi.org/10.1145/3282286.3282299","url":null,"abstract":"Feature extraction is the most important part of the whole object recognition and target detection system. Convolutional Networks have evolved to the state-of-the-art technique for computer vision tasks owing to the predominant feature extraction capability. However, the working process of Convolutional Networks is invisible, which makes it difficult to optimize the model. To evaluate a Convolutional Network, we introduce a novel way to project the activities back to the input pixel space, revealing what input pattern originally caused a specific activation in the feature maps. Using this visualization technique, we take the feature extraction of sunflower seed image containing an impurity as an example, and attempt to change the architecture of traditional Convolutional Networks in order to extract better specific features for target images. After a series of improvements, we got a new Convolutional Network which is more conducive to the target images feature extraction and the number of parameters is less than before, which is conducive to the transplantation of the small system. Our model can be docking the state-of-the-art recognition networks according to different application scenarios, so as to structure a complete automatic recognition system.","PeriodicalId":324982,"journal":{"name":"Proceedings of the 2nd International Conference on Graphics and Signal Processing","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132231249","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Face recognition: Sparse Representation vs. Deep Learning 人脸识别:稀疏表示与深度学习

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282291

Neamah H. Alskeini, Kien Nguyen Thanh, V. Chandran, W. Boles

{"title":"Face recognition: Sparse Representation vs. Deep Learning","authors":"Neamah H. Alskeini, Kien Nguyen Thanh, V. Chandran, W. Boles","doi":"10.1145/3282286.3282291","DOIUrl":"https://doi.org/10.1145/3282286.3282291","url":null,"abstract":"The pose, illumination and facial expression discrepancies between two face images are the key challenges in face recognition. The deep Convolutional Neural Networks (CNNs) and the fast Sparse Representation-based Classification (SRC) have achieved promising results in face recognition. However, CNNs require large databases and extremely expensive computations to overcome other algorithms. In this paper, we propose a novel SRC-based algorithm using test input image sets and training sub-databases, and compare its performance with CNNs. Histograms of Oriented Gradients (HOG) descriptors are used to define a new technique, named Training Image Modification (TIM), which provides image training sets with large variations of faces. The proposed algorithm divides the image training set into a number of sub-databases to address the dimensionality problem, and uses a test input image set to extract a signature from each sub-database using SRC. Each signature contains the same number of images as the test image set, although these may belong to different subjects. Considering all the sub-databases sequentially, the algorithm uses the signature of each sub-database to compute the number of images belonging to each subject. The signature that produces the Maximum Number of Images (MNI) of the same subject will have captured this subject for identification. YouTube Celebrity (YTC) and Multi-PIE databases are used in this work to evaluate the efficacy of the proposed method, which achieves high recognition rates. For relatively small databases, the proposed method is simple, scalable and stable, and it results in good face recognition rate under large face variations, as demonstrated by comparison with CNNs.","PeriodicalId":324982,"journal":{"name":"Proceedings of the 2nd International Conference on Graphics and Signal Processing","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122844352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 5

Phenotypic Parameter Extraction System for Crops Based on Supervoxel Segmentation 基于超体素分割的作物表型参数提取系统

Proceedings of the 2nd International Conference on Graphics and Signal Processing Pub Date : 2018-10-06 DOI: 10.1145/3282286.3282294

Jiafeng Zheng, Geng Liu, Xiangpeng Liu

引用次数: 1