检索结果-内蒙古大学图书馆

7th International Conference on Image, Vision and Computing, ICIVC 2022

作者： Liu, Song Li, Haiwei Zhang, Geng Hu, Bingliang Chen, Junyu Xi'an Institute of Optics and Precision Mechanics Chinese Academy of Sciences Key Laboratory of Spectral Imaging Technology of Cas Xi'an China University of Chinese Academy of Sciences Beijing China

ISBN: (数字)9781665467346

ISBN: (纸本)9781665467346

In recent years, using multispectral remote sensing data to change detection has been a hot issue. However, the limited number of spectral segments cannot accurately classify ground objects, which will affect the performance of change detection. To overcome this problem, in this paper, we propose a multispectral image change detection method based on a hyperspectral reconstruction network. The original multispectral data is used to generate hyperspectral data with richer spectral information, and then the generated data is used for change detection. The whole network consists of a reconstruction network and a change detection network. Based on the input multispectral data, a small amount of hyperspectral data is used to assist the training. In addition, we design a weighted loss to balance the two subnetworks in the training stage. In the test stage, the network can automatically recover hyperspectral data from multispectral data for change detection. The experimental results on two data sets show that the hyperspectral reconstruction method can improve the accuracy of multispectral images. © 2022 IEEE.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Modulated Pulse Characteristics Based on Nonlinear Frequency Upconversion 8

Modulated Pulse Characteristics Based on Nonlinear Frequency...

引用

8th Symposium on Novel Photoelectronic Detection technology and Applications

作者： Li, Yan Han, Yongcheng Gu, Guohua He, Weiji Chen, Qian Jiangsu Key Laboratory of Spectral Imaging and Intelligence Sense Nanjing University of Science and Technology Nanjing210094 China

ISBN: (数字)9781510653122

ISBN: (纸本)9781510653115

We report on numerical analysis of the conversion efficiency under different pumping conditions in the nonlinear frequency upconversion system, the influence of initial frequency chirp is also discussed in detail. Our results demonstrate that the wave peak is proportional to the amplitude parameter U. In addition, the positive chirp is conducive to spectrum broadening, while the negative chirp does the opposite. These numerical investigation will be meaningful for further improving the conversion efficiency and provide some reference for quantum communications and quantum manipulation in the single-photon frequency upconversion process. © 2022 SPIE

关键词： Conversion efficiency

来源：评论

学校读者我要写书评

暂无评论

DC-KD: double-constraint knowledge distillation for optical satellite imagery object detection based on YOLOX model 4

DC-KD: double-constraint knowledge distillation for optical ...

引用

4th International Conference on Machine Learning and Computer Application, ICMLCA 2023

作者： Yang, Hongbo Qiu, Shi Feng, Xiangpeng Key Laboratory of Spectral Imaging Technology CAS Xi'an Institute of Optics and Precision Mechanics Chinese Academy of Science Xi'an China University of Chinese Academy of Sciences Beijing China

ISBN: (数字)9781510680265

ISBN: (纸本)9781510680258

Object detection is an important application of optical satellite remote sensing imagery interpretation. Since the objects of interest, such as aircraft, ships, and vehicles, are small in size with obscure contour and texture, it is difficult for object detection in satellite images. The spatial resolution of aerial images is higher than satellite images, and the object detection model can achieve higher precision. Knowledge distillation has been validated as an effective technique by learning the common features of aerial and satellite images to improve the precision of object detection in satellite images. It means that a teacher model pre-trained on aerial image datasets guides the training of a compact student model on satellite image datasets. However, there are data distribution differences between aerial images and satellite images. The distribution differences may cause the teacher model to give guidance signals that deviate from the ground truth, thus leading to sub-optimization of the student model. In this paper, we proposed a new distillation scheme, termed DC-KD, which updates the teacher model using the predictions of the teacher model that are inconsistent with the ground truth, and the rest are used to guide the training of the student model. We achieved a 3.88% mAP50 improvement on the xView dataset based on the YOLOX-S model. © 2024 SPIE.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Exploiting optical degrees of freedom for information multiplexing in diffractive neural networks

引用

Light(Science & Applications) 2022年第8期11卷 1639-1642页

作者： Chao Zuo Qian Chen Smart Computational Imaging Laboratory(SCILab) School of Electronic and Optical EngineeringNanjing University of Science and Technology210094NanjingJiangsu ProvinceChina Smart Computational Imaging Research Institute(SCIRI)of Nanjing University of Science and Technology 210019NanjingJiangsu ProvinceChina Jiangsu Key Laboratory of Spectral Imaging&Intelligent Sense 210094NanjingJiangsu ProvinceChina

Exploiting internal degrees of freedom of light, such as polarization, provides efficient ways to scale the capacity of optical diffractive computing, which may ultimately lead to high-throughput, multifunctional all-... 详细信息

关键词： degrees optical networks

来源：评论

学校读者我要写书评

暂无评论

An Optimized Feature extraction Algorithm Based on SuperPoint

An Optimized Feature extraction Algorithm Based on SuperPoin...

引用

2022 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2022

作者： Bian, Xiaobei He, Zhen Gong, Ziqing Ren, Kan Jiangsu Key Laboratory of Spectral Imaging and Intelligent Sense Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9781510662407

Feature extraction and matching of remote sensing images is becoming increasingly important with a wide range of applications. It matches and superimposes images obtained from the same scene at different times, different sensors, and different angles, and maps the optimal to the target image. CNN-based algorithms have shown superior expressiveness compared to traditional methods in almost all fields with image. This paper optimises the network based on SuperPoint by replacing convolution with a depth-separable convolution which has smaller number of parameters, and replacing the conv block with a spindle-type Inverted Residuals block composed of dimension expansion, depth-separable convolution and Dimension reduction. The network depth is fine-tuned to ensure accuracy. The model is trained on the RSSCN7 remote sensing dataset. Compared with other traditional algorithms in a cross-sectional manner with the combination of SuperGlue, the optimized algorithm shows the superior comprehensive performance. © 2023 SPIE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Remote Sensing Scene Classification with Multi-task Learning 7th

Remote Sensing Scene Classification with Multi-task Learning

引用

7th China High-resolution Earth Observation Conference, CHREOC 2020

作者： Gong, Tengfei Zheng, Xiangtao Lu, Xiaoqiang Key Laboratory of Spectral Imaging Technology CAS Xi’an Institute of Optics and Precision Mechanics Chinese Academy of Sciences Xi’an China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9789811657344

Remote sensing scene classification is a fundamental task for understanding the high-resolution remote sensing images and has achieved great progress thanks to the deep neural networks. However, the deep neural networks heavily depend on the large-scale training samples, while sufficient manually annotated data is often unavailable. In this paper, a Multi-Task Scene Classification Network (MTSCN) is proposed to address the label consumption dilemma by using annotated information from other related tasks. The MTSCN simultaneously conducts multiple tasks so that scene classification task can leverage information shared by other tasks. The MTSCN consists of a shared branch and multiple task-specific branches with each for one task. First, the shared branch extracts shared features for all tasks. Then, the task-specific branch adapts shared features into task-specific features to finish the classification task. Such architecture allows complementary information shared among tasks as well as avoids task interference by selecting specific information for a different task. The method is evaluated on two multi-task learning scenarios with respect to accuracy and the number of network parameters. The results show that the proposed method is efficient not only in terms of data but also in terms of network parameters. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Remote sensing

来源：评论

学校读者我要写书评

暂无评论

Enhancing Audio-Visual Spiking Neural Networks through Semantic-Alignment and Cross-Modal Residual Learning

arXiv

引用

arXiv 2025年

作者： He, Xiang Zhao, Dongcheng Dong, Yiting Shen, Guobin Yang, Xin Zeng, Yi Brain-inspired Cognitive Intelligence Lab Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing100049 China Brain-inspired Cognitive Intelligence Lab Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Future Technology University of Chinese Academy of Sciences Beijing100049 China CAS Key Laboratory of Molecular Imaging Institute of Automation Chinese Academy of Sciences Beijing100190 China Center for Long-term Artificial Intelligence Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Brain Cognition and Brain-inspired Intelligence Technology Chinese Academy of Sciences Shanghai200031 China

Humans interpret and perceive the world by integrating sensory information from multiple modalities, such as vision and hearing. Spiking Neural Networks (SNNs), as brain-inspired computational models, exhibit unique advantages in emulating the brain’s information processing mechanisms. However, existing SNN models primarily focus on unimodal processing and lack efficient cross-modal information fusion, thereby limiting their effectiveness in real-world multimodal scenarios. To address this challenge, we propose a semantic-alignment cross-modal residual learning (S-CMRL) framework, a Transformer-based multimodal SNN architecture designed for effective audiovisual integration. S-CMRL leverages a spatiotemporal spiking attention mechanism to extract complementary features across modalities, and incorporates a cross-modal residual learning strategy to enhance feature integration. Additionally, a semantic alignment optimization mechanism is introduced to align cross-modal features within a shared semantic space, improving their consistency and complementarity. Extensive experiments on three benchmark datasets CREMA-D, UrbanSound8K-AV, and MNISTDVS-NTIDIGITS demonstrate that S-CMRL significantly outperforms existing multimodal SNN methods, achieving the state-of-the-art performance. The code is publicly available at https://***/Brain-Cog-Lab/S-CMRL. Copyright © 2025, The Authors. All rights reserved.

关键词： Information fusion

来源：评论

学校读者我要写书评

暂无评论

Effect of light source coherence on reconstructed targets by speckles-correlation algorithm 9

Effect of light source coherence on reconstructed targets by...

引用

9th Symposium on Novel Photoelectronic Detection technology and Applications

作者： Miao, Jinye Shi, Yingjie Bai, Lianfa Guo, Enlai Han, Jing Jiangsu Key Laboratory of Spectral Imaging and Intelligent Sense Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9781510664432

Based on the optical memory effect, many speckle-correlation methods have been proposed for imaging through scattering media. However, most of these methods directly take the laser with a rotating diffuser as a pseudo-thermal source without discussing its physical basis. To explore the effect of different optical coherence on object reconstruction, the principle that rotating diffuser can weaken the coherence is discussed and validated by experiments in this paper. Specifically, the degree of coherence is changed by controlling both the speed of the diffuser and the size of the illumination spot. And the speckle patterns collected under these conditions are reconstructed by speckle-correlation algorithms (SCA). Through theoretical derivation and experiments, it is well demonstrated that the low coherence facilitates the recovery of the object. In addition, it is found that the stacked ground glass moving along a single direction can replace the rotating diffuser to achieve light source decoherence. The effect of different optical coherence on scattering imaging is explored in this paper, which provides a theoretical basis for selecting the light source in speckle-correlation imaging. © 2023 SPIE.

关键词： Light sources

来源：评论

学校读者我要写书评

暂无评论

A simple tree structure for dynamic programming based on real scene

A simple tree structure for dynamic programming based on rea...

引用

2013 International Conference on Advanced Information Engineering and Education Science, ICAIEES 2013

作者： Pang, Xing Zhang, Yi Bai, Lianfa Han, Jing Jiangsu Key Laboratory of Spectral Imaging and Intelligent Sense Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9789078677949

A new stereo matching approach based on first order invariant moment is proposed. First order invariant moment is utilized in our algorithm to transform the left and right images from the input space to the feature space, and the improved a simple tree structure dynamic programming ba-sed on adaptive parameters is introduced to evaluate the disparity map to reduce the mismatching points on discontinuous and low texture regions. To testify the applicability of the proposed algorithm, it is applied to gray images captured in the real situation, and the results of experiments show that the proposed algorithm can effectively improves the quality of disparity map on interference areas. © 2013. The authors - Published by Atlantis Press.

关键词： Dynamic programming

来源：评论

学校读者我要写书评

暂无评论

Research on Infrared Small Target Segmentation Algorithm

Research on Infrared Small Target Segmentation Algorithm

引用

2022 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2022

作者： Chen, Zhuo Bian, Xiaobei Ren, Kan Wan, Minjie Jiangsu Key Laboratory of Spectral Imaging and Intelligent Sense Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9781510662407

The task of segmenting small infrared targets, which have few pixels and weak features, has been a difficult problem in the field of small target image processing. Small targets exist not only in general images, but also widely in UAV cameras, communication base station cameras, rescue cameras and vehicle cameras. The study of small target segmentation algorithms is very important for analyzing and utilizing these images, and has important applications in security, transportation, and rescue. Traditional small target segmentation algorithms are able to segment objects with simple target contour edges and large differences in signal strength. The traditional algorithm often has high false detection rate and missed detection rate when facing several targets with weak signal strength, and does not perform well in complex scenes. In this paper, we introduce an infrared small target segmentation scheme facing multiple types and numbers of targets. We also produce an infrared UAV and pedestrian dataset for validation. © 2023 SPIE.

关键词： Cameras

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：