检索结果-内蒙古大学图书馆

11th International Conference on Bio-inspired Computing – Theories and Applications, BIC-TA 2016

作者： Tian, Xiaolin Zhao, Sujie Jiao, Licheng Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education International Research Center of Intelligent Perception and Computation International Collaboration Joint Lab in Intelligent Perception and Computation Xidian University XianShaanxi710071 China

ISBN: (纸本)9789811036101

In this paper, we propose a novel visual tracking method based on ensemble learning using logistic regression model. We adopt logistic regression to achieve ensemble classifier to deal with object tracking problem. By using fast computable features, our approach learns the appearance of the target during tracking. And thus, the proposed method is able to adapt online to target appearance changes and its surrounding background. Moreover, ensemble learning converts rough rules of thumb into highly accurate prediction rule. Experimental results show that our method outperforms relative trackers. © Springer Nature Singapore Pte Ltd. 2016.

关键词： Regression analysis

来源：评论

学校读者我要写书评

暂无评论

intelligent multi-user detection using an artificial immune system

引用

Science in China(Series F) 2009年第12期52卷 2342-2353页

作者： GONG MaoGuo, JIAO LiCheng, MA WenPing & MA JingJing key laboratory of intelligent perception and image understanding of ministry of education of China Institute of intelligent Information Processing, Xidian University, Xi’an 710071, China 1. Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China Xidian University Xi’an 710071 China 2. Institute of Intelligent Information Processing Xidian University Xi’an 710071 China

Artificial immune systems （AIS） are a kind of new computational intelligence methods which draw inspiration from the human immune system. In this study, we introduce an AIS-based optimization algorithm, called clonal selection algorithm, to solve the multi-user detection problem in code-division multipleaccess communications system based on the maximum-likelihood decision rule. Through proportional cloning, hypermutation, clonal selection and clonal death, the new method performs a greedy search which reproduces individuals and selects their improved maturated progenies after the affinity maturation process. Theoretical analysis indicates that the clonal selection algorithm is suitable for solving the multi-user detection problem. Computer simulations show that the proposed approach outperforms some other approaches including two genetic algorithm-based detectors and the matched filters detector, and has the ability to find the most likely combinations.

关键词： artificial immune systems clonal selection multi-user detection code-division multiple-access genetic algorithm

来源：评论

学校读者我要写书评

暂无评论

Building a Stereo and Wide-View Hybrid RGB/FIR Imaging System for Autonomous Vehicle

引用

IEEE Sensors Journal 2022年第2期22卷 1638-1651页

作者： Zhang, Yigong Sun, Zezhou Wu, Banghe Yang, Jian Kong, Hui Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education Jiangsu Key Laboratory of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China University of Macau Taipa China

In this work, a novel paradigm is proposed to build a hybrid RGB/FIR camera, which outputs pixel-wisely aligned and temporally synchronized RGB and FIR frames simultaneously. Compared with the existing schemes, we innovate in several ways to make it compact, practical, and extendable for real-world applications. First, we propose a new structure layout design of the hybrid camera, and we also design a removable adjuster to detach the hybrid camera from the base platform. For RGB and FIR image registration, we have achieved a pixel-wise spatial registration of the FIR and RGB frames by a two-step alignment, where we propose a ghost-effect removal scheme to remove the ghost effect from the rolling-shutter-based FIR camera. Finally, multiple hybrid cameras are extended to a hybrid camera array, through which we can obtain wide-view spatially aligned FIR, RGB and disparity images simultaneously. The experimental results demonstrate the accuracy of the hybrid camera and the effectiveness of the designed stereo wide-view hybrid camera system. © 2001-2012 IEEE.

关键词： Cameras

来源：评论

学校读者我要写书评

暂无评论

Tracking algorithm based on dual residual network and kernel correlation filters 2nd

Tracking algorithm based on dual residual network and kernel...

引用

2nd CCF International Conference on Artificial Intelligence, CCF-ICAI 2019

作者： Tian, Xiaolin Li, Fang Jiao, Licheng Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education International Research Center of Intelligent Perception and Computation International Collaboration Joint Lab in Intelligent Perception and Computation Xidian University XianShaanxi Province710071 China

ISBN: (纸本)9789813292970

Visual target tracking is a target detection task for a period of time. During this period, the tracking target will undergo significant appearance changes due to deformation, sudden movement, complex background and occlusion. These changes make visual target tracking challenging. In this paper, a target tracking algorithm based on dual residual neural network and kernel correlation filters is proposed, which mainly solves the problems of inaccurate tracking. This method combines depth residual neural network with kernel correlation filter tracking algorithm. The template matches the depth residual feature to determine the location of the target and the kernel correlation filter based on residual feature is designed to detect the target, finally, the template matching result and kernel correlation filtering result are fused to determine the location of the target. The experimental results on a large-scale standard data set show that the proposed algorithm has the advantages of high accuracy. Compared with the previous algorithm, the proposed algorithm has good performance in tracking visual deformation, occlusion and fuzzy video objects. © 2019, Springer Nature Singapore Pte Ltd.

关键词： Target tracking

来源：评论

学校读者我要写书评

暂无评论

A Dual Residual Network with Channel Attention for image Restoration 16th

A Dual Residual Network with Channel Attention for Image Res...

引用

Workshops held at the 16th European Conference on Computer Vision, ECCV 2020

作者： Nie, Shichao Ma, Chengconghui Chen, Dafan Yin, Shuting Wang, Haoran Jiao, LiCheng Liu, Fang Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education International Research Center for Intelligent Perception and Computation Joint International Research Laboratory of Intelligent Perception and Computation School of Artificial Intelligence Xidian University Xi’an Shaanxi710071 China

ISBN: (纸本)9783030682378

Deep learning models have achieved significant performance on image restoration task. However, restoring the images with complicated and combined degradation types still remains a challenge. For this purpose, we proposed a dual residual network with channel attention (DRANet) to address complicated degradation in the real world. We further exploit the potential of encoder-decoder structure. To fuse feature more efficiently, we adopt the channel attention module with skip-connections. To better process low- and high-level information, we introduce the dual residual connection into the network architecture. And we explore the effect of multi-level connection to image restoration. Experimental results demonstrate the superiority of our proposed approach over state-of-the-art methods on the UDC T-OLED dataset. © 2020, Springer Nature Switzerland AG.

关键词： image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Fine-grained image categorization with fisher vector 11th

Fine-grained image categorization with fisher vector

引用

11th International Conference on Bio-inspired Computing – Theories and Applications, BIC-TA 2016

作者： Tian, Xiaolin Ding, Xin Jiao, Licheng Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education International Research Center of Intelligent Perception and Computation International Collaboration Joint Lab in Intelligent Perception and Computation Xidian University Xi’anShaanxi Province710071 China

ISBN: (纸本)9789811036132

Fine-grained image categorization is a categorization task, where classifying objects should be the same basic-level class and have similar shape or visual appearances. Generally, the bag-of-words (BoW) model is popular in image categorization. However, in BoW model, the feature quantization for image representation is also a lossy process, which severely limits the descriptive power of the image representation. Fisher vectors employ soft assignments and reduce information loss due to quantization by calculating the gradient for each parameter separately, which have been shown to outperform other global representations on most benchmark datasets. In this paper, the acquired template is represented by Fisher Vector (FV). Combing FV with improved spatial pyramid matching (SPM) respectively, we use an approach, i.e., FV+SPM, to obtain feature representation. Experimental results show that our method outperforms state-of-the-art categorization approaches on the Caltech-UCSD Birds dataset. © Springer Nature Singapore Pte Ltd. 2016.

关键词： Template matching

来源：评论

学校读者我要写书评

暂无评论

Unified spatial masking for just-noticeable difference estimation

Unified spatial masking for just-noticeable difference estim...

引用

Asia-Pacific Signal and Information Processing Association Annual Summit and Conference 2011, APSIPA ASC 2011

作者： Wu, Jinjian Qi, Fei Shi, Guangming Key Laboratory of Intelligent Perception and Image Understanding Ministry of Education of China China School of Electronic Engineering Xidian University Xi'an Shaanxi 710071 China

In this paper, we introduce a unified spatial masking function for the estimation of just-noticeable difference (JND). Conventional models estimate several parts independently, and then combine these parts to get the JND. In this work, we treat the spatial masking effect as a nonlinear transformation of the luminance adaptation. To model the transformation, we measure the deviation of image contents from the ideal patterns to establish luminance adaptation rules. Considering both luminance difference and structural regularity, we derive a nonlinear spatial masking function by modulating luminance adaptation with the deviation coefficients. The masking function deduces an accurate estimation of the JND. Experiments demonstrate the validity of the proposed framework.

关键词： Luminance

来源：评论

学校读者我要写书评

暂无评论

Accelerating Deep Convolutional Neural Network Inference Based on OpenCL 5th

Accelerating Deep Convolutional Neural Network Inference Bas...

引用

5th IFIP TC 12 International Conference on Intelligence Science, ICIS 2022

作者： Wu, Yong Zhu, Huming Zhang, Lingyun Hou, Biao Jiao, Licheng Xidian-Wuhu Research Institute Wuhu China Key Laboratory of Intelligent Perception and Image Understanding Ministry of Education Xidian University Xi’an710071 China

ISBN: (纸本)9783031149023

In recent years, in order to facilitate the efficient application of deep convolutional neural networks, it has become increasingly important to accelerate the inference stage of deep convolutional neural networks. But with the development of numerous heterogeneous computing devices, today’s popular deep learning inference tools only support specific devices, so they cannot effectively utilize different GPU devices to accelerate DNN inference. To address this issue, we propose an OpenCL-based parallel deep convolutional neural network inference algorithms. Firstly, we design and implement parallel kernel code using OpenCL to accelerate depthwise separable convolution, and implement parallel matrix multiplication combined with clBLAS to accelerate traditional convolution. Meanwhile, we design OpenCL parallel kernel codes for other operations in the inference stage of deep convolutional neural networks. Secondly, we further improve the inference performance by means of kernel fusion and increasing the workload per core. Finally, MobileNet v1 network and the 21-layer residual network based on OpenCL are run on AMD Radeon Vega Frontier GPU and Nvidia GeForce GTX 1070 GPU. Compared to the Caffe implementation, 40.16x, 1.67x speedups are achieved on the AMD GPU and 14.95x, 1.11x speedups are achieved on the Nvidia GPU. © 2022, IFIP International Federation for Information Processing.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

DELVING INTO SEMANTIC SCALE IMBALANCE

DELVING INTO SEMANTIC SCALE IMBALANCE

引用

11th International Conference on Learning Representations, ICLR 2023

作者： Ma, Yanbiao Jiao, Licheng Liu, Fang Li, Yuxin Yang, Shuyuan Liu, Xu Key Laboratory of Intelligent Perception and Image Understanding of the Ministry of Education School of Artificial Intelligence Xidian University Xi'an710071 China

Model bias triggered by long-tailed data has been widely studied. However, measure based on the number of samples cannot explicate three phenomena simultaneously: (1) Given enough data, the classification performance gain is marginal with additional samples. (2) Classification performance decays precipitously as the number of training samples decreases when there is insufficient data. (3) Model trained on sample-balanced datasets still has different biases for different classes. In this work, we define and quantify the semantic scale of classes, which is used to measure the feature diversity of classes. It is exciting to find experimentally that there is a marginal effect of semantic scale, which perfectly describes the first two phenomena. Further, the quantitative measurement of semantic scale imbalance is proposed, which can accurately reflect model bias on multiple datasets, even on sample-balanced data, revealing a novel perspective for the study of class imbalance. Due to the prevalence of semantic scale imbalance, we propose semantic-scale-balanced learning, including a general loss improvement scheme and a dynamic re-weighting training framework that overcomes the challenge of calculating semantic scales in real-time during iterations. Comprehensive experiments show that dynamic semantic-scale-balanced learning consistently enables the model to perform superiorly on large-scale long-tailed and non-long-tailed natural and medical datasets, which is a good starting point for mitigating the prevalent but unnoticed model bias. Furthermore, we look ahead to future challenges in the field. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Weakly Supervised Land Cover Classification Method for Large-Scale Multi-Resolution Labeled Satellite images Data Sets

Weakly Supervised Land Cover Classification Method for Large...

引用

2020 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2020

作者： Yin, Shuting Chen, Dafan Ma, Chengconghui Lian, Yanchao Jiao, Licheng Liu, Fang Joint International Research Laboratory of Intelligent Perception and Computation School of Artificial Intelligence Xidian University Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education International Research Center for Intelligent Perception and Computation Shaanxi Province Xian710071 China

The global land cover map obtained by satellite is of vital importance to record land information. Despite the large amount of data acquired, rare annotations are available. Weakly-supervised learning methods can help to get use of extensive existing data resources and reduce the cost of manual labeling, which is of great significance for large-scale land cover mapping. Considering that the low-resolution labels are inaccurate and inexact and that high-resolution labels are rare, which is likely to cause training overfitting, we propose a method that can effectively apply multi-resolution labels, combining machine learning and deep learning. First, high-resolution labels of the validation set are used to train DeepLabv3+ (DLv3) and Random Forest (RF), which are applied to predict the training set then. The intersection of the prediction results is then used to supervise the training of DLv3 as pseudo-labels together with the high-resolution labels. We also added strategies such as model fusion, data augmentation and weighted training to improve classification accuracy. In the end, our method achieves the average accuracy (AA) of 0.609, ranking 3rd in Track 2 of the IEEE Data Fusion Contest (DFC) of 2020. © 2020 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：