A novel content based image retrieval （CBIR） algorithmusing relevant feedback is presented. The proposed frameworkhas three major contributions： a novel feature descriptor calledcolor spectral histogram （CSH） to ...A novel content based image retrieval （CBIR） algorithmusing relevant feedback is presented. The proposed frameworkhas three major contributions： a novel feature descriptor calledcolor spectral histogram （CSH） to measure the similarity betweenimages; two-dimensional matrix based indexing approach proposedfor short-term learning （STL）; and long-term learning （LTL）.In general, image similarities are measured from feature representationwhich includes color quantization, texture, color, shapeand edges. However, CSH can describe the image feature onlywith the histogram. Typically the image retrieval process starts byfinding the similarity between the query image and the imagesin the database; the major computation involved here is that theselection of top ranking images requires a sorting algorithm to beemployed at least with the lower bound of O（n log n）. A 2D matrixbased indexing of images can enormously reduce the searchtime in STL. The same structure is used for LTL with an aim toreduce the amount of log to be maintained. The performance ofthe proposed framework is analyzed and compared with the existingapproaches, the quantified results indicates that the proposedfeature descriptor is more effectual than the existing feature descriptorsthat were originally developed for CBIR. In terms of STL,the proposed 2D matrix based indexing minimizes the computationeffort for retrieving similar images and for LTL, the proposed algorithmtakes minimum log information than the existing approaches.展开更多
This paper presents a novel efficient semantic image classification algorithm for high-level feature indexing of high-dimension image database. Experiments show that the algorithm performs well. The size of the train ...This paper presents a novel efficient semantic image classification algorithm for high-level feature indexing of high-dimension image database. Experiments show that the algorithm performs well. The size of the train set and the test set is 7 537 and 5 000 respectively. Based on this theory, another ground is built with 12,000 images, which are divided into three classes: city, landscape and person, the total result of the classifications is 88.92%, meanwhile, some preliminary results are presented for image understanding based on semantic image classification and low level features. The groundtruth for the experiments is built with the images from Corel database, photos and some famous face databases.展开更多
The problem considered in this paper is how to detect the degree of similarity in the content of digital images useful in image retrieval,i.e.,to what extent is the content of a query image similar to content of other...The problem considered in this paper is how to detect the degree of similarity in the content of digital images useful in image retrieval,i.e.,to what extent is the content of a query image similar to content of other images.The solution to this problem results from the detection of subsets that are rough sets contained in covers of digital images determined by perceptual tolerance relations(PTRs).Such relations are defined within the context of perceptual representative spaces that hearken back to work by J.H.Poincare on representative spaces as models of physical continua.Classes determined by a PTR provide content useful in content-based image retrieval(CBIR).In addition,tolerance classes provide a means of determining when subsets of image covers are tolerance rough sets(TRSs).It is the nearness of TRSs present in image tolerance spaces that provide a promising approach to CBIR,especially in cases such as satellite images or aircraft identification where there are subtle differences between pairs of digital images,making it difficult to quantify the similarities between such images.The contribution of this article is the introduction of the nearness of tolerance rough sets as an effective means of measuring digital image similarities and,as a significant consequence,successfully carrying out CBIR.展开更多
To realize content-hased retrieval of large image databases, it is required to develop an efficient index and retrieval scheme. This paper proposes an index algorithm of clustering called CMA, which supports fast retr...To realize content-hased retrieval of large image databases, it is required to develop an efficient index and retrieval scheme. This paper proposes an index algorithm of clustering called CMA, which supports fast retrieval of large image databases. CMA takes advantages of k-means and self-adaptive algorithms. It is simple and works without any user interactions. There are two main stages in this algorithm. In the first stage, it classifies images in a database into several clusters, and automatically gets the necessary parameters for the next stage-k-means iteration. The CMA algorithm is tested on a large database of more than ten thousand images and compare it with k-means algorithm. Experimental results show that this algorithm is effective in both precision and retrieval time.展开更多
知识检索增强技术通过引入外部知识库有效缓解了大语言模型(large language models,LLM)的幻觉问题与知识滞后性,成为提升领域任务性能的关键范式。本文针对电网设备缺陷等级识别任务中标注样本缺乏、知识利用率低与可解释性不足的问题...知识检索增强技术通过引入外部知识库有效缓解了大语言模型(large language models,LLM)的幻觉问题与知识滞后性,成为提升领域任务性能的关键范式。本文针对电网设备缺陷等级识别任务中标注样本缺乏、知识利用率低与可解释性不足的问题,提出了一种零样本的知识增强的大语言模型协同推理框架。构建了层次化树状知识库,设计语义最相关的二阶段检索算法提升知识获取效率,并创新性地融合大语言模型先验知识与检索知识进行多阶段推理验证。该方法在218例测试数据上取得54.17%的分类准确率,较无知识检索方法提升了14.26%,同时通过思维链提示生成可验证的解释文本。此外,该方法为零样本,不需要标注数据进行训练。实验结果表明,本方法有效发挥了专业领域知识与通用知识的协同作用,为电力设备缺陷自动检测提供了准确与可解释的解决方案。展开更多
文摘A novel content based image retrieval （CBIR） algorithmusing relevant feedback is presented. The proposed frameworkhas three major contributions： a novel feature descriptor calledcolor spectral histogram （CSH） to measure the similarity betweenimages; two-dimensional matrix based indexing approach proposedfor short-term learning （STL）; and long-term learning （LTL）.In general, image similarities are measured from feature representationwhich includes color quantization, texture, color, shapeand edges. However, CSH can describe the image feature onlywith the histogram. Typically the image retrieval process starts byfinding the similarity between the query image and the imagesin the database; the major computation involved here is that theselection of top ranking images requires a sorting algorithm to beemployed at least with the lower bound of O（n log n）. A 2D matrixbased indexing of images can enormously reduce the searchtime in STL. The same structure is used for LTL with an aim toreduce the amount of log to be maintained. The performance ofthe proposed framework is analyzed and compared with the existingapproaches, the quantified results indicates that the proposedfeature descriptor is more effectual than the existing feature descriptorsthat were originally developed for CBIR. In terms of STL,the proposed 2D matrix based indexing minimizes the computationeffort for retrieving similar images and for LTL, the proposed algorithmtakes minimum log information than the existing approaches.
文摘This paper presents a novel efficient semantic image classification algorithm for high-level feature indexing of high-dimension image database. Experiments show that the algorithm performs well. The size of the train set and the test set is 7 537 and 5 000 respectively. Based on this theory, another ground is built with 12,000 images, which are divided into three classes: city, landscape and person, the total result of the classifications is 88.92%, meanwhile, some preliminary results are presented for image understanding based on semantic image classification and low level features. The groundtruth for the experiments is built with the images from Corel database, photos and some famous face databases.
基金supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) research grants 194376 and 185986Manitoba Centre of Excellence Fund(MCEF) grant and Canadian Network Centre of Excellence(NCE) and Canadian Arthritis Network(CAN) grant SRI-BIO-05.
文摘The problem considered in this paper is how to detect the degree of similarity in the content of digital images useful in image retrieval,i.e.,to what extent is the content of a query image similar to content of other images.The solution to this problem results from the detection of subsets that are rough sets contained in covers of digital images determined by perceptual tolerance relations(PTRs).Such relations are defined within the context of perceptual representative spaces that hearken back to work by J.H.Poincare on representative spaces as models of physical continua.Classes determined by a PTR provide content useful in content-based image retrieval(CBIR).In addition,tolerance classes provide a means of determining when subsets of image covers are tolerance rough sets(TRSs).It is the nearness of TRSs present in image tolerance spaces that provide a promising approach to CBIR,especially in cases such as satellite images or aircraft identification where there are subtle differences between pairs of digital images,making it difficult to quantify the similarities between such images.The contribution of this article is the introduction of the nearness of tolerance rough sets as an effective means of measuring digital image similarities and,as a significant consequence,successfully carrying out CBIR.
基金This project was supported by National High Tech Foundation of 863 (2001AA115123)
文摘To realize content-hased retrieval of large image databases, it is required to develop an efficient index and retrieval scheme. This paper proposes an index algorithm of clustering called CMA, which supports fast retrieval of large image databases. CMA takes advantages of k-means and self-adaptive algorithms. It is simple and works without any user interactions. There are two main stages in this algorithm. In the first stage, it classifies images in a database into several clusters, and automatically gets the necessary parameters for the next stage-k-means iteration. The CMA algorithm is tested on a large database of more than ten thousand images and compare it with k-means algorithm. Experimental results show that this algorithm is effective in both precision and retrieval time.
文摘知识检索增强技术通过引入外部知识库有效缓解了大语言模型(large language models,LLM)的幻觉问题与知识滞后性,成为提升领域任务性能的关键范式。本文针对电网设备缺陷等级识别任务中标注样本缺乏、知识利用率低与可解释性不足的问题,提出了一种零样本的知识增强的大语言模型协同推理框架。构建了层次化树状知识库,设计语义最相关的二阶段检索算法提升知识获取效率,并创新性地融合大语言模型先验知识与检索知识进行多阶段推理验证。该方法在218例测试数据上取得54.17%的分类准确率,较无知识检索方法提升了14.26%,同时通过思维链提示生成可验证的解释文本。此外,该方法为零样本,不需要标注数据进行训练。实验结果表明,本方法有效发挥了专业领域知识与通用知识的协同作用,为电力设备缺陷自动检测提供了准确与可解释的解决方案。
