Retrieval Technology Research Articles

Synergies and Challenges in the Integration of Cloud Computing and Deep Learning: Current Status, Interconnectedness, and Future Directions

This article reviews the status and recent developments in the integration of cloud computing and deep learning, as well as the interrelationship between these two technologies. The paper explores the intersection of cloud computing and deep learning in addressing cybersecurity challenges. Amidst the rapid expansion of the worldwide public cloud services market, the vulnerability to cyber-attacks and breaches in data management is on the rise. Different intrusion detection systems use different deep learning techniques to improve the effectiveness of intrusion detection in cloud computing environments. Additionally, the use of encryption technology and the corresponding deep learning retrieval technology further improves the security of cloud data. Moreover, the paper deeply studies how the scheduling mechanism of deep reinforcement learning can optimize the performance of cloud services by efficiently allocating resources and solving the problem of slow cloud service speed. It also derives the optimal energy strategy through deep neural networks to address the energy consumption challenges in cloud computing data centers. This article also reviews the five emerging architectures of cloud computing and explores the role of deep learning within these frameworks. Finally, it analyzes some of the challenges facing the future of cloud computing and deep learning, including the security and confidentiality of cloud computing, as well as low latency and high throughput optimization in the field of deep learning. In summary, this article provides insight into current trends, challenges, and future prospects for the evolving integration between cloud computing and deep learning.

Personalized music teaching service recommendation based on sensor and information retrieval technology

In order to fully utilize the existing resources of music information, people have found that music information retrieval technology has important research significance for personalized music teaching. Overall, music information retrieval technology is still in the experimental exploration stage and lacks technical and practical systems. In this context, this article proposes recommendation algorithms that can effectively adjust the content of information services based on user preferences. However, recommendation algorithms are still in an immature and complex stage, and there are still issues with their accuracy. In response to the issue of recommendation accuracy, this article utilizes information retrieval technology to optimize recommendation algorithms. Combine these two algorithms for recommendation. The main steps that affect recommendation results in recommendation algorithms include similarity calculation, nearest neighbor selection method, and score prediction method calculation. The experimental results indicate that the proposed algorithm can effectively solve the existing problems of users, and can also accurately retrieve the content of interest, greatly improving the diversity of the proposed content. We use utilizing algorithmic learning to create a personalized music teaching system that enables students to independently advance towards their self-development goals, ultimately achieving the goal of promoting the harmonious development of students' personalities.

Design And Development of Sketch Based Image Retrieval Using Deep Learning

In this cutting edge, the common wrong doing rate is expanding day-by-day and to manage up with this the criminal divisions as well ought to discover ways in which would speed up the by and large preparation and offer assistance in bringing one to justice. In response to rising crime rates, law enforcement agencies are turning to advanced algorithms capable of matching freehand sketches with images in databases. These algorithms, utilizing sophisticated feature extraction techniques and deep learning models, significantly enhance identification accuracy. By leveraging Sketch based image retrieval technology, investigations are expedited, leading to quicker suspect apprehensions and resolution of criminal cases. This results in improved public safety and justice outcomes, as well as more efficient law enforcement practices overall.

A Survey on Content Retrieval on the Decentralised Web

The control, governance, and management of the web have become increasingly centralised, resulting in security, privacy, and censorship concerns. Decentralised initiatives have emerged to address these issues, beginning with decentralised file systems. These systems have gained popularity, with major platforms serving millions of content requests daily. Complementing the file systems are decentralised search engines and name-registry infrastructures, together forming the basis of a decentralised web . This survey article analyses research trends and emerging technologies for content retrieval on the decentralised web, encompassing both academic literature and industrial projects. Several challenges hinder the realisation of a fully decentralised web. Achieving comparable performance to centralised systems without compromising decentralisation is a key challenge. Hybrid infrastructures, blending centralised components with verifiability mechanisms, show promise to improve decentralised initiatives. While decentralised file systems have seen more mature deployments, they still face challenges such as usability, performance, privacy, and content moderation. Integrating these systems with decentralised name-registries offers a potential for improved usability with human-readable and persistent names for content. Further research is needed to address security concerns in decentralised name-registries and enhance governance and crypto-economic incentive mechanisms.

An Investigation of Distance Measures for Development of Effective Content Based Tumor Image Retrieval System

Background/Objectives: The MRI has proven to be extremely effective in detecting tumors, with millions of images created each day throughout the world. To find similar images from a vast collection, Content-Based Tumor Image Retrieval (CBTIR) technology has been used to analysis the medical image. In the traditional retrieval methods, retrieving a similar image from the large database is crucial task. To overcome this issue we developed deep learning based retrieval method. Methods: This research offers a retrieval approach based on predefined ResNet models for quick and accurate image retrieval. We tested various prominent ResNet models with different distance similarity metrics, and the best option was determined by this work. Findings: After the various evaluation of ResNet models with varied distance measures on the CE-MRI data set, ResNet50 model applied with Hamming distance yields 99.33% of retrieval precision. Novelty: This work used predefined ResNet models with the combination of Distance similarity metrics to achieve more accurate results on medical image retrieval compared to the other conventional methods. Keywords: Content Based Image Retrieval, Tumor Retrieval, Hamming Distance, Euclidean Distance, Minkowski

Bidirectional phase retrieval: Protecting the imaging of cells and tissues from interference of noise on the carrier

Quantitative phase imaging plays a crucial role in biological imaging, with many phase retrieval technologies being particularly important. However, cells and tissues must be placed on certain carriers during imaging, such as glass slides, cell culture dishes, and cell culture boxes. Therefore, current unidirectional phase retrieval technologies are not suitable for biplane samples, causing the phase of the object to always be coupled with the phase of noise. As a solution, this study proposes bidirectional phase retrieval to independently analyze the phase of both planes in the sample. Considering that most biological samples can be treated as two planes: one representing the object and the other serving as noise. By decoupling the phase of the two planes, this approach protects biplane biological samples from interference caused by the noise plane, leading to more accurate phase retrieval. This innovation holds significant value for the advancement of biological imaging.

Visual Language – Let the Product Say What You Want

Visual Language is a multitasking on-line system focusing on e-commerce, which involves in generating accurate product descriptions for sellers and providing convenient product retrieval service for customers. To achieve this goal, the system adopts image description technology and multi-modal retrieval technology. By utilizing cross-modal generation technique, we could help sellers on rapid uploading products and customers on rapid retrieval, which could improve the experience of both sellers and customers.

Application and Research of Media Information Retrieval Technology in Student Work Management Systems

The development of the internet has promoted the improvement of media information retrieval technology, which has also been applied to the work management of college students. This paper mainly adopts media information retrieval technology, studies the function principle of this technology, and establishes a student work management system based on media information retrieval technology. Then, this system is applied to the student management of different grades, and the problems existing in the student work management of each grade are compared and analyzed, and the corresponding improvement measures are put forward. The results indicate that the student work management system can improve the efficiency of student work management among colleges and universities in different grades.

Integrated Design of Building Environment Based on Image Segmentation and Retrieval Technology

Existing models still exhibit a deficiency in capturing more detailed contextual information when processing architectural images. This paper introduces a model for architectural image segmentation and retrieval based on an image segmentation network. Primarily, spatial attention is incorporated into the U-Net segmentation network to enhance the extraction of image features. Subsequently, a dual-path attention mechanism is integrated into the U-Net backbone network, facilitating the seamless integration of information across different spaces and scales. Experimental results showcase the superior performance of the proposed model on the test set, with average dice coefficient, accuracy, and recall reaching 94.67%, 95.61%, and 97.88%, respectively, outperforming comparative models. The proposed model can enhance the U-Net network's capability to identify targets within feature maps. The amalgamation of image segmentation networks and attention mechanisms in artificial intelligence technology enables precise segmentation and retrieval of architectural images.

Mass Spectrometry-Based Method to Measure Aflatoxin B1 DNA Adducts in Formalin-Fixed Paraffin-Embedded Tissues.

Aflatoxin B1 (AFB1) is a potent human liver carcinogen produced by certain molds, particularly Aspergillus flavus and Aspergillus parasiticus, which contaminate peanuts, corn, rice, cottonseed, and ground and tree nuts, principally in warm and humid climates. AFB1 undergoes bioactivation in the liver to produce AFB1-exo-8,9-epoxide, which forms the covalently bound cationic AFB1-N7-guanine (AFB1-N7-Gua) DNA adduct. This adduct is unstable and undergoes base-catalyzed opening of the guanine imidazolium ring to form two ring-opened diastereomeric 8,9-dihydro-8-(2,6-diamino-4-oxo-3,4-dihydropyrimid-5-yl-formamido)-9-hydroxy-aflatoxin B1 (AFB1-FapyGua) adducts. The AFB1 formamidopyrimidine (Fapy) adducts induce G → T transversion mutations and are likely responsible for the carcinogenic effects of AFB1. Quantitative liquid chromatography-mass spectrometry (LC-MS) methods have shown that AFB1-N7-Gua is eliminated in rodent and human urine, whereas ring-opened AFB1-FapyGua adducts persist in rodent liver. However, fresh frozen biopsy tissues are seldom available for biomonitoring AFB1 DNA adducts in humans, impeding research advances in this potent liver carcinogen. In contrast, formalin-fixed paraffin-embedded (FFPE) specimens used for histopathological analysis are often accessible for molecular studies. However, ensuring nucleic acid quality presents a challenge due to incomplete reversal of formalin-mediated DNA cross-links, which can preclude accurate quantitative measurements of DNA adducts. In this study, employing ion trap or high-resolution accurate Orbitrap mass spectrometry, we demonstrate that ring-opened AFB1-FapyGua adducts formed in AFB1-exposed newborn mice are stable to the formalin fixation and DNA de-cross-linking retrieval processes. The AFB1-FapyGua adducts can be detected at levels comparable to those in a match of fresh frozen liver. Orbitrap MS2 measurements can detect AFB1-FapyGua at a quantification limit of 4.0 adducts per 108 bases when only 0.8 μg of DNA is assayed on the column. Thus, our breakthrough DNA retrieval technology can be adapted to screen for AFB1 DNA adducts in FFPE human liver specimens from cohorts at risk of this potent liver carcinogen.

Multitarget scattering imaging based on Nonnegative Matrix Factorization beyond the optical memory effect

Speckle correlation technique is limited by the optical memory effect (OME) range, and aliasing phenomenon occurs in multitarget information beyond the range. Here, we propose a multitarget scattering imaging method based on Nonnegative Matrix Factorization (NMF) beyond the OME, which can separate and interpret the single target information from the aliasing information. Based on the study of the multitarget aliasing speckle autocorrelation model, we apply NMF to unmix the multitarget speckle autocorrelation beyond the OME, and combine with phase retrieval (PR) technology to recover the target. Our approach requires only single-shot speckle information without prior information to realize wide-field multitarget imaging. This work provides potential value for biomedical imaging and astronomical observation, and helps to broaden the application scope of speckle correlation technique.

Development of key technologies for audio retrieval and vocal teaching system based on sensor networks

Traditional audio processing technology often faces the challenge of efficiency and accuracy in the face of large-scale audio data and complex sound content. The purpose of this study is to develop the key technology of audio retrieval and vocal music teaching system based on sensor network, aiming at the rich audio data and real-time monitoring function provided by sensor network. The sensors suitable for audio retrieval and vocal music teaching were selected and rationally arranged in the designated position to obtain accurate and comprehensive audio data. The algorithm is also optimized and improved to improve the performance and stability of the system. The accuracy and efficiency of audio retrieval can be effectively improved by using sensor networks. The vocal music teaching system based on sensor networks can provide comprehensive and accurate sound evaluation and guidance. Good results have been obtained in the practical application of this method, which proves its feasibility and effectiveness in the field of audio retrieval and vocal music teaching.

A Case Study of Millennial English Teachers’ Awareness of Digital Media in EFL Classrooms

Digital literacy embodies the competence in effectively employing technology for information retrieval, appraisal, creation, and transmission. This case study not only explores the digital literacy of millennial English teachers but uniquely unveils the seamless integration of a diverse array of digital media into their EFL classrooms. The questionnaire served as a means to gauge the extent of the millennial English teachers understood the notion of digital literacy. Meanwhile, to understand how they use digital technology resources in their instructional design and pedagogical strategies, this study deployed observation, interview, and document analysis. Ten English teachers from Cianjur who belong to the millennial generation enlisted in the study. The findings underscored that the participants adeptly utilized an array of digital technology tools in their day-to-day professional undertakings. They have demonstrated proficiency with digital teaching tools and have seamlessly integrated them into their EFL classroom. Nonetheless, there is a discernible need for further support in advancing their digital literacy skills, and further, collaborative environment to explore innovative pedagogical approaches. Also, a sufficient amount of time for professional development is necessary, as are policies that promote the effective use of mobile devices in the classroom. Furthermore, it is imperative to stimulate the teachers’ experimentation with emerging technologies and elevate their digital literacy levels. Therefore, it is advised that government and policymakers proffer substantial, reliable, and easily accessible technical support tailored to the needs of millennial English teachers to continue their digital literacy growth.

Development Path of Teaching Method Innovation in Art Colleges and Universities under the Background of Internet

Abstract The informatization teaching method is a big challenge to the traditional form of teaching in art college education. This paper studies the construction of an informatization teaching method combining multimedia technology and multimodal media information retrieval technology. In view of the demand and development conception of teaching methods in art colleges and universities under the background of the Internet, it examines the willingness of teachers of art colleges and universities to teach informatization as well as the degree of satisfaction of their teaching ability. It optimizes the retrieval of learning resources based on the multi-category classification method of ELM classifiers and proposes an integrated mode of informatization teaching in art disciplines. According to the factors affecting the informatization teaching ability of art teachers, teaching design and teacher implementation are carried out in a case-by-case manner to analyze and summarize the program effects of the informatization teaching approach. The correlation coefficient between teachers’ willingness to adopt informative teaching methods and their satisfaction with their information teaching ability is 0.614, which is moderately correlated. After the implementation of the teaching program, the total score of the teacher’s informatization teaching ability reaches 80.052, which improves the level of informatization teaching compared to the level of informatization in pre-preparation. The informatization teaching method’s program design can enhance the teacher’s informatization teaching level, which can have an impact on the innovative development of the teaching method in art colleges and universities through practical experience.

Practical Analysis of Digital Technology in Physical Education and Training in the Context of Multimedia Era

Abstract In the multimedia era, the application of digital technology in physical education and training is becoming increasingly important. This paper analyzes how digital technology can improve biological education teaching effectiveness and students’ motivation and independent learning ability. The research methodology uses digital physical education teaching and curriculum integration, motion retrieval technology, and motion editing and scoring technology. It is found that the physical education teaching mode using digital technology is significantly better than the traditional method, in which the advantage of the class using digital technology in the mastery of physical education skills is obvious. The Analysis of final grades shows that the number of experimental courses with more than 90 points is three times as many as that of the control class. The average grade, learning motivation, and independent learning ability are higher than in traditional teaching. The specific data show that the mean value of understanding motivation in the experimental class is 26.47, and that in the control class is 24.87 (P=0.0031), which shows that the digital technology significantly improves students’ motivation to learn. The application of digital technology in physical education can dramatically improve the teaching effect, enhance students’ learning motivation and independent learning ability, an important development direction of modern physical education.

A Classification and Retrieval System for Learning Resources of MOOC

In this information age, massive open online courses (MOOCs) have become an integral component of modern education. These courses encompass a wide range of resources, such as videos, audios, texts, and other forms. An accompanying question is how to effectively organize, classify, and retrieve these resources. However, currently available classification and retrieval methods are mostly based on text retrieval technologies. As a result, multi-modal resources such as videos and audios are often ignored or incorrectly classified. Furthermore, more current methods exhibit low efficiency when processing the vast amount of data in MOOCs. To address and solve these issues, this study focuses on the extraction and fusion of multi-modal features of MOOC resources. It proposes an efficient classification and retrieval method based on 3D convolution, aiming to offer a more accurate and efficient approach for classifying and retrieving MOOC resources.

Electronic Health File System based on Fingerprint Sensor Technology

In medical sector, the use of computers and information technology helps in diagnosis, especially the computer-based patient record (CPR). As the traditional method of storing patient files is not feasible to obtain the health file, especially those who are in critical conditions, such as accidents, loss of consciousness and memory at times, due to the consumption of a lot of time and effort in the process of manual searching in traditional files. In order to help health care centers quickly identify patient personal data and make accurate decisions, an automated electronic system for patient health files based on a fingerprint sensor was built using special methods and technologies such as Content-Based Image Retrieval (CBIR) system, which is a retrieval technology based on the similarity of the images has the potential to facilitate identification and diagnosis of the patient's condition when he or she is unconscious. This study aims to integrate different technologies and discuss issues related to the electronic health file system based on the patient's fingerprint data, the use of (fingerprint sensor) technology, linking to the patient file data system, storage in fingerprint databases, and health data for each. sick.; To help critical cases, in addition to applying this technology and employing it in the ambulance and diagnostic system in order to improve the level of medical services and reach the desired results. The results showed that the electronic health profile was successfully built, designed and implemented for its intended purpose, the system and the destinations that connect users to the system as well as the fingerprint sensor. Finally, the system is efficient and gives reliable results with up to 100% tested accuracy based on the states it was stored.

Secure Image Retrieval and Sharing Technologies for Digital Inclusive Finance: Methods and Applications

Secure Image Retrieval and Sharing Technologies for Digital Inclusive Finance: Methods and Applications

A New Mode of Interaction to Explore All-sky Maps in Multiple Wavelengths

As the Museums and Interactivity-oriented Science Centres of the world are reinventing themselves more as creators of culture than just conservators of culture, they are getting keener to explore how various data retrieval and HCI technologies can symbiotically fit into the methods of museology. Museums, since their advent, have used spatial and temporal scanning to present stories that evolve either over time or through space. But these have always remained static timelines that are limited in content, lack interactive exploration, and miss any metaphorical interaction that allows one to have the experience of physically scanning a space or a timeline. In this work, we report how a new HCI technology was used to retrieve all-sky map data and present the multi-wavelength view of the universe to the uninitiated visitor having no skill to navigate through sophisticated astronomy instruments. We describe an intuitive interface that takes no training and learning for the visitors to operate. They can start analyzing multi-wavelength all-sky maps of our galaxy in no time – exactly as professional astronomers do. Stressing upon the ease of usability, reliability of data, and fail-safe operation, visitors are given access to additional layers of information that can be presented in myriad options chosen by the visitor. They can compare thin slices of galactic images in various wavelengths and look for essential radiation sources that are otherwise not apparent in the naked eye or simple optical observations. The public exhibit was based on duplex data communication between a computer and a PIC microprocessor controller using IEEE 1284 parallel port in ECP mode thus making it a generic technology for interactive displays.

The Application of Artificial Intelligence Technology in Public Library Information Retrieval

With the continuous innovation and development of social science and technology, artificial intelligence technology has been the focus of attention and research in more and more industries. Public library, as an important place in the society to provide mass education for the masses, should fully seize the development opportunity with artificial intelligence technology as the core idea in the operation and development, carry out intelligent, automatic and digital reform and innovation of the library, and optimize the links of information retrieval, book borrowing, information service and access to the library. Through Internet technology, artificial intelligence technology and other advanced technologies, the whole science and technology content of public libraries and public service quality is improved so as to meet the specific needs of the public for the use of public education resources. Based on this, this paper first briefly describes the basic concepts of human intelligence technology and information retrieval technology, and then studies the specific application of artificial intelligence technology in public library information retrieval link and related service applications, hoping to provide some reference for professionals.