检索结果-内蒙古大学图书馆

LWIR hyperspectral image classification based on a temperature-emissivity residual network and conditional random field model

引用

INTERNATIONAL JOURNAL OF REMOTE SENSING 2022年第10期43卷 3744-3768页

作者： Cao, Liqin He, Jiani Gao, Lyuzhou Zhong, Yanfei Hu, Xin Li, Zhijiang Wuhan Univ Graph Commun & Printing & Packaging Res Ctr Wuhan 430072 Peoples R China Zhijiang Lab Hangzhou Peoples R China Wuhan Univ State Key Lab Informat Engn Surveying Mapping & R Wuhan 430072 Peoples R China Wuhan Univ Sch Informat Management Wuhan Peoples R China

Long-wave infrared (LWIR) hyperspectral images can provide temperature and emissivity information, and have unique advantages in classification tasks, especially in the fields of minerals, vegetation, and man-made materials. However, LWIR image acquisition is greatly affected by the atmosphere, which results in obvious spectral noise. Meanwhile, field-based and aerial thermal infrared hyperspectral images are faced with problems such as spatial heterogeneity and salt-and-pepper noise, which results in the traditional spectral-based and object-oriented classification methods performing poorly in the classification task. In this paper, using the emissivity spectra and land surface temperature inversed from LWIR hyperspectral images, a classification framework based on a temperature-emissivity residual network and conditional random field model (TERN-CRF) is proposed. In this method, the residual network is designed to extract and fuse in-depth spectral and local spatial features, and the conditional random field (CRF) model incorporates the spatial-contextual information by the use of the temperature information, to improve the problem of voids and isolated regions in the classification map. To validate the performance of the proposed method, three LWIR datasets acquired by the Hyper-Cam sensor were used in the experiments: an aerial dataset and two field-based datasets of minerals and leaves. The experimental classification results obtained using the LWIR datasets confirm the performance of the proposed TERN-CRF classification method, which is shown to be superior to the traditional hyperspectral image classification methods and a CNN-based method. Furthermore, the CRF post-processing part can effectively alleviate the isolated regions in the classification map.

关键词： Fourier-transformed long-wave infrared hyperspectral imagery spectral emissivity conditional random field model deep learning classification

来源：评论

学校读者我要写书评

暂无评论

Boundary Recognition Method for Preposition Phrases in Modern Chinese based on Top-N Algorithm

引用

JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 2025年第2期41卷 419-434页

作者： Zhang, Hui-fang Sichuan Univ Sch Literature & Media Jinjiang Coll Pengshan 620860 Peoples R China

In modern Chinese sentences, the structure and usage of prepositional phrases are very flexible and diverse, which makes it difficult to mine the topic of prepositional phrases in modern Chinese, and leads to the decline of the accuracy of prepositional phrase boundary recognition. Therefore, Athe paper studies the boundary recognition method for preposition phrases in modern Chinese based on Top-N algorithm. Firstly, a prepositional phrase corpus is constructed through a prepositional usage attribute dictionary, a prepositional usage rule base and a prepositional usage corpus. Then, based on the corpus of prepositional phrases, the topic of prepositional phrases in modern Chinese is mined by LDA model, and the modern Chinese documents are segmented. conditional random field modelAis used to extract prepositional phrase features from segmented modern Chinese documents. Finally, the recognition of prepositional phrase boundaries in modern Chinese is realized based on local model weighted fusion Top-N algorithm. The experimental results show that in the three types of prepositional phrase environments, the highest accuracy of prepositional phrase boundary recognition achieved by the research method is 99.6%, and the highest F-value is 99.7%. On the surface, this method can accurately identify the boundaries of modern Chinese prepositional phrases and is practical.

关键词： Top-N algorithm modern Chinese preposition phrase boundary recognition method LDA model conditional random field model

来源：评论

学校读者我要写书评

暂无评论

A FULLY CONNECTED CHANGE DETECTION METHOD OF SAR IMAGES FUSING ORIGINAL IMAGE FEATURES AND CHANGE DETECTION RESULTS 5

A FULLY CONNECTED CHANGE DETECTION METHOD OF SAR IMAGES FUSI...

引用

5th International-Society-for-Photogrammetry-and-Remote-Sensing (ISPRS) Geospatial Week (GSW)

作者： Sun, Zhentao Duan, Fuzhou Guan, Hongliang Yang, Fan Wang, Yanhui Zhao, Wenji Capital Normal Univ Coll Resources Environm & Tourism 105 West Third Ring North Rd Beijing Peoples R China Minist Educ Space Informat Technol Engn Res Ctr 105 West Third Ring North Rd Beijing Peoples R China

The primary strategy to eliminate the effect of scatter noise in synthetic aperture radar (SAR) imagery is usually through filtering or combining neighborhood information. However, both approaches to reducing noise reduce the detection accuracy of change edges with similar characteristics to scatter noise points. Considering the above problems, this letter proposes a post-processing method that applies a fully connected conditional random field theoretical model to fuse the original image information with the initial change detection results. The method first takes the original image information and the initial change detection results as a priori conditions. Secondly, the global spatial information in the original image and the label values in the initial change detection results are fully considered when detecting the changed and unchanged pixels to establish a fully connected relationship between all the pixels and find the label distribution probability of each pixel under the condition of noise suppression, and finally obtain better change detection results. The experimental results on the real SAR dataset confirm the proposed method's effectiveness, robustness, and efficiency.

关键词： SAR Change Detection Full connection conditional random field model Post-processing

来源：评论

学校读者我要写书评

暂无评论

Chinese Word Segmentation Based on Deep Learning 2018

Chinese Word Segmentation Based on Deep Learning

引用

10th International Conference on Machine Learning and Computing (ICMLC)

作者： Wang, Mengge Li, Xiaoge Wei, Zheng Zhi, Shuting Wang, Haoyue Xian Univ Posts & Telecommun Sch Comp Xian Shaanxi Peoples R China Shaanxi Prov Sci & Technol Resource Coordinat Ctr Xian Shaanxi Peoples R China

ISBN: (纸本)9781450363532

Chinese word segmentation is a fundamental task in the field of Chinese Natural Language Processing. In this paper, we propose a series of neural network architectures by combining Long Short-Term Memory Neural Network (LSTM) with conditional random field (CRF). Firstly, we use the Skip-Gram model to obtain character embeddings from a raw corpus (without word delimiters). Then, we treat these embeddings as inputs of LSTM model to compute their context representation vectors. Finally, these representation vectors are applied to the CRF layer for Chinese word segmentation. Experiments on the corpus of the 2014 people's daily and Chinese Computing (NLPCC2015) Weibo texts segmentation task show that our models outperforms the previous machine learning methods.

关键词： Chinese word segmentation character embedding Long Short-Term Memory model conditional random field model

来源：评论

学校读者我要写书评

暂无评论

Research on ARIA of Affective Recommendation Index Algorithm Based on Topic Selection of Publications

Research on ARIA of Affective Recommendation Index Algorithm...

引用

International Conference on Big Data and Artificial Intelligence (BDAI)

作者： Sun, Huayan Li, Ye-Li Zi, Yun-Fei Han, Xu Beijing Inst Graph Commun Sch Informat Engn Beijing 102600 Peoples R China

ISBN: (纸本)9781538661369

At present, the publishing organization's publication selection planning method lags behind, making China's publishing industry economic benefits greatly reduced. This paper presents a new affective analysis method-Affective Recommendation index (ARIA) based on the method of fine-grained affective analysis, which combines conditional random field model (CRF) and affective weighting method to analyze the evaluation information of publication, Targeted selection of the evaluation object and its attributes analysis of the reader's emotional inclination, taking full account of the emotion-free words but the influence of the invisible affective sentences on the emotion, the author analyzes the emotion of the book review with affective words and analyses the emotion of the book review, and makes a weighted calculation of the emotional inclination, and concludes that the readers are positive to the publication, The negative evaluation quantity, through "The introduction weight value-weighted sum-Calculates the ratio" the algorithm step, obtains the Affective recommendation index. To match the readers ' interest by the index of affective recommendation, more accurate grasp of the book publishing market and user needs, reduce the backlog of books inventory, achieve more accurate publication, personalized recommendations, increase the economic benefits of publishing industry.

关键词： Topics of publication fine-grained affective analysis conditional random field model affective weighting calculation evaluation object attribute affective recommendation index

来源：评论

学校读者我要写书评

暂无评论

Trajectory Data Association Based MultiObject Tracking 29

Trajectory Data Association Based MultiObject Tracking

引用

第29届中国控制与决策会议

作者： Dongliang Jin Songhao Zhu Yanyun Cheng School of Automatic Nanjing University of Posts and Telecommunications

ISBN: (纸本)9781509046584

A multitarget tracking method based on data association and trajectory evaluation is proposed in this ***,the data association and trajectory evaluation are integrated into the same conditional random field model,which is transformed to obtain the minimum energy ***,from the time,if two of the same track label appear with a physical space,the symbiotic label cost is used to constrain the correlation;from the space,the pairwise energy term between the different observation targets is introduced to prevent the occurrence of false data ***,in the discrete continuous optimization process of the energy,this paper uses the improved a-expansion algorithm and the gradient descent method to solve the minimum energy in non convex and non sub module *** experimental results of PETS2009/2010 benchmark and TUD-Stadtmitte video sequence database show that the proposed algorithm in this paper is superior to the current advanced level of multitarget tracking technology.

关键词： multitarget tracking conditional random field model pairwise energy term multilabel costs discrete and continuous optimization

来源：评论

学校读者我要写书评

暂无评论

Semantic labelling of SAR images with conditional random fields on region adjacency graph

引用

IET RADAR SONAR AND NAVIGATION 2011年第8期5卷 835-841页

作者： Yang, W. Chen, L. Dai, D. Xia, G. -S. Wuhan Univ Sch Elect Informat Wuhan 430079 Peoples R China TELECOM ParisTech CNRS LTCI F-75013 Paris France

Scene segmentation and semantic labelling are important for analysing and understanding synthetic aperture radar (SAR) images. In this study, the authors propose an effective and efficient labelling method for SAR images with conditional random fields on a region adjacency graph (CRF-RAG). More precisely, for an SAR image, a region adjacency graph (RAG) representation is firstly built on an initially over-segmentation of the image. Subsequently, a conditional random field (CRF) model is established over the RAG instead of over pixels. To train and infer the CRF-RAG model, a fast max-margin training strategy and the graph cut optimisation method are finally employed. As the CRF model is based on RAG, the computation complexity of the model can be reduced significantly. Compared to the Markov random field (MRF) model on RAG, the proposed CRF-RAG model is more efficient to incorporate different measures of SAR images, such as scattering intensity, texture and image context, into a unified model. Experiments on the TerraSAR-X imagery achieve promising results with modest computation cost, which validates the generality and flexibility of the proposed method.

关键词： scattering intensity semantic labelling Optimisation techniques radar imaging synthetic aperture radar images image segmentation TerraSAR-X imagery Optical, image and video signal processing optimisation image context image representation Radar equipment, systems and applications Combinatorial mathematics image over-segmentation MRF model SAR images graph cut optimisation method RAG representation computation cost image texture synthetic aperture radar max-margin training strategy computation complexity CRF-RAG model random processes conditional random fields region adjacency graph representation CRF model labelling method computational complexity Markov random field model graph theory conditional random field model scene segmentation Markov processes

来源：评论

学校读者我要写书评

暂无评论

Soft tagging of overlapping high confidence gene mention variants for cross-species full-text gene normalization

引用

BMC BIOINFORMATICS 2011年第s8期12卷 S6-S6页

作者： Kuo, Cheng-Ju Ling, Maurice H. T. Hsu, Chun-Nan Acad Sinica Inst Informat Sci Taipei 115 Taiwan Singapore Polytech Sch Chem & Life Sci Dover Singapore Univ Melbourne Dept Zool Parkville Vic 3052 Australia Univ So Calif Inst Informat Sci Marina Del Rey CA 90292 USA

Background: Previously, gene normalization (GN) systems are mostly focused on disambiguation using contextual information. An effective gene mention tagger is deemed unnecessary because the subsequent steps will filter out false positives and high recall is sufficient. However, unlike similar tasks in the past BioCreative challenges, the BioCreative III GN task is particularly challenging because it is not species-specific. Required to process full-length articles, an ineffective gene mention tagger may produce a huge number of ambiguous false positives that overwhelm subsequent filtering steps while still missing many true positives. Results: We present our GN system participated in the BioCreative III GN task. Our system applies a typical 2-stage approach to GN but features a soft tagging gene mention tagger that generates a set of overlapping gene mention variants with a nearly perfect recall. The overlapping gene mention variants increase the chance of precise match in the dictionary and alleviate the need of disambiguation. Our GN system achieved a precision of 0.9 (F-score 0.63) on the BioCreative III GN test corpus with the silver annotation of 507 articles. Its TAP-k scores are competitive to the best results among all participants. Conclusions: We show that despite the lack of clever disambiguation in our gene normalization system, effective soft tagging of gene mention variants can indeed contribute to performance in cross-species and full-text gene normalization.

关键词： conditional random field Gene Mention conditional random field model Dictionary Match NCBI Taxonomy Database

来源：评论

学校读者我要写书评

暂无评论

Pretreatment for Speech Machine Translation

Pretreatment for Speech Machine Translation

引用

2nd International Conference on Computational Collective Intelligence: Technologies and Applications

作者： Zhang, Xiaofei Feng, Chong Huang, Heyan Chinese Acad Sci Res Ctr Comp & Language Informat Engn Beijing 100097 Peoples R China Beijing Inst Technol Sch Comp Technol Beijing 100081 Peoples R China

ISBN: (纸本)9783642167317

There are many meaningless modal particles and dittographes in natural spoken language, furthermore ASR (automatic speech recognition) often has some recognition errors and the ASR results have no punctuations. And thus the translation would be rather poor if the ASR results are directly translated by MT (machine translation). Therefore, it is necessary to transform the abnormal ASR results into normative texts to fit machine translation. In this paper, a pretreatment approach which based on conditional random field model was introduced to delete the meaningless modal particles and dittographes, correct the recognition errors, and punctuated the ASR results before machine translation. Experiments show that the MT BLEU of 0.2497 is obtained, that improved by 18.4% over the MT baseline without pretreatment.

关键词： Speech machine translation automatic speech recognition conditional random field model pretreatment

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：