检索结果-内蒙古大学图书馆

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Salim, Devano Hofker Susanto, Jose Juan Manalu, Sonya R. Shiddiqi, Hafizh Ash School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia School of Computer Science Bina Nusantara University Mobile Application & Technology Program Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350390025

The digital landscape is undergoing profound changes, with visual search technology like Google Lens emerging as a pivotal force shaping our interaction with information. This research paper delves into Google Lens, a cutting-edge tool redefining how we engage with visual data. We comprehensively examine its functionalities, underlying principles, and real-world applications, particularly in education and library services. Through meticulous analysis, we evaluate its efficacy in addressing practical challenges and assess its accuracy, efficiency, and breadth of capabilities compared to other visual search technologies. By exploring existing research and the intricacies of Google Lens, this paper aims to provide a holistic understanding of this technology and its potential impact on how we access and interact with information in the real world. Finally, we explore the future of technology by contemplating the prospective roles of Google Lens across various industries and its potential for shaping the future of information access. © 2024 IEEE.

关键词： future of technology google lens visual search technology

来源：评论

学校读者我要写书评

暂无评论

AI-Powered Mobile Application for Image-Based Food Ingredient Detection and Recipe Generation 9

AI-Powered Mobile Application for Image-Based Food Ingredien...

引用

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Rosaline, Thessalonica Maria Luisse Suciawan, Calvin Anacia Ng, William Achmad, Said Moniaga, Jurike V School of Computer Science Bina Nusantara University Mobile Application & Technology Program Computer Science Department Jakarta11480 Indonesia School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350390025

This research explores the critical impact of dietary habits on health, addressing the lack of awareness regarding nutritional content that leads to issues such as malnutrition. The proposed mobile application utilizes the YOLOv8 algorithm with a deep convolutional neural network (CNN) architecture, deployed on the Roboflow server, a local MySQL database server, and ChatGPT. The architectural framework of this application is integrated through the NodeJs server via an API system embedded within the Android application. Despite this approach, the study highlights the need for further refinement in the accuracy of the ingredients' object detection model. The refined YOLOv8 model, trained on an extensive dataset of 24,583 images across 13 food ingredient classes, outperforms the initial YOLOv8 model trained on 9,867 images. Improvements are notable, particularly in addressing misclassifications of ingredients such as beef, pork, onion, and egg in the initial model. However, the YOLOv8 model can still be improved by introducing more ingredient variations and expanding the training dataset. This application can detect multiple food ingredients from an image and evaluate their nutritional value, specifically in calorie count. It also provides tailored recipes to help users meet their daily nutritional requirements, offering a practical solution for users to achieve their dietary goals. This research connects computer vision and nutrition, providing insights for developing practical solutions in this domain. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Quantitative analysis of sign language translation using artificial neural network model 9th

Quantitative analysis of sign language translation using art...

引用

9th International Conference on computer science and Computational Intelligence, ICCSCI 2024

作者： Wijaya, Fendy Dahendra, Leonardo Purwanto, Eko Setyo Ario, Muhamad Keenan Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia Mobile Application & Technology Program Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia

Sign language is one of the technique to support communication with deaf and speech impaired people. Nowadays, human needs become more complex, so are the needs of people with those disabilities. Therefore, with the sophistication of modern technology in the field of computer science, it is necessary to have a sign language translation system which is capable to convert human gestures into words or spelling letters in a natural human language, especially Indonesian language and SIBI sign language. This research is conducted to analyze the capabilities of an Artificial Neural Network (ANN) system to translate SIBI sign language into Indonesian language. The analysis is performed quantitatively from the experiment results to gain a descriptive insights for the generated models. By creating two datasets containing alphabets and words in SIBI sign language, two models were generated to predict alphabets and words, respectively. The result shows that ANN model could effectively and efficiently perform a sign language translation with accuracy of alphabet prediction reached 96.15% and words prediction reached 99.45% in duration of prediction without exceeding 0.15 seconds. Regarding to the results, it can be concluded that ANN model quantitatively suitable to be implemented as an Indonesian sign language translator system. © 2024 The Authors.

关键词： Neural machine translation

来源：评论

学校读者我要写书评

暂无评论

Parallelization of LSTM-GRU Architectures for Multivariate Prediction of Stock Prices 9

Parallelization of LSTM-GRU Architectures for Multivariate P...

引用

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Tarigan, Gabriel Asael Hermawan, Eric Girsang, Abba Suganda Bina Nusantara University BINUS Graduate Program - Master of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350390025

Anticipating stock market movements is challenging, as hasty prediction risks large financial losses. Stock price prediction is complex due to the high volatility and many influencing external factors. Recurrent Neural Network (RNN) models are made to learn time series data making it ideal for stock prediction based on records. Well-known RNN architectures for forecasting problems are Long-Short Term Memory (LSTM) and Gated Recurrent Unit (GRU). However, past studies only explored standalone models and have not generated convincing performance results. By integrating the efficiency of GRU to effectively handle shorter data sequences with LSTM's ability to capture long-term dependencies, our model aims to improve stock price forecasting accuracy by collaborating the strength between both architectures. The proposed model was used to forecast multivariate stock prices and evaluated using Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). We found that parallelization of LSTM-GRU offers better prediction, up to 9.1% improvement compared to sequential LSTM-GRU and Bidirectional LSTM (Bi-LSTM). © 2024 IEEE.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Implementation of Gamification to Help Teenagers Prevent Diseases 9

Implementation of Gamification to Help Teenagers Prevent Dis...

引用

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Haryanto, Michael Djaja, Fabian Putera Chrisandy, Michael Achmad, Said Moniaga, Jurike V Bina Nusantara University Mobile Application & Technology Program Computer Science Department School of Computer Science Jakarta11480 Indonesia Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia

ISBN: (纸本)9798350390025

Adolescence is an important part of growing up to become an adult;keeping adolescent physique healthy is very important because poor dietary choices and a sedentary lifestyle will lead to diseases that they shouldn't have experienced yet. That is why teenagers need to be engaged to do more health- promoting behaviors, one way is by implementing gamification. This paper seeks to study which kind of gamification elements can be implemented from past research and whether the results were effective or not to prove whether gamification is a valid and effective strategy. At the same time, we are creating a gamified prototype for the mobile platform. The methodology starts with a systematic review of PubMed, MDPI, and Google Scholar to get a general idea of which gamification elements are the most implemented. Then a prototype will be made with the most used gamified elements from the systematic review. And then participants ranging from 10 - 19 years old will test the prototype and will give their opinions via a survey, From this, it is proven by our own gamified prototype survey results that using gamification as a way to engage teenagers is a fun and engaging way to do more health-promoting behaviors with a mostly positive feedbacks with the average rating of 8.33 out of 10 given from the participants. © 2024 IEEE.

关键词： Gamification

来源：评论

学校读者我要写书评

暂无评论

Enhancing User Interface and Experience in Skin Type Classification and Skincare Product Recommendation System for Young Adults 9

Enhancing User Interface and Experience in Skin Type Classif...

引用

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Himawan, Felicia Soerjadi, Vania Paramitha Handoyo, Yonathan Achmad, Said Moniaga, Jurike V School of Computer Science Bina Nusantara University Mobile Application & Technology Program Computer Science Department Jakarta11480 Indonesia School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350390025

The escalating interest of skincare among Indonesian young adults reflects the rising trend in the skincare industry. However, this heightened interest also brings several challenges, like the lack of knowledge and confusion when selecting the right skincare products for our skin type. This underscores the need for an effective and efficient skin type classification and product recommendation system. This paper addresses this challenge by focusing on enhancing a skincare application's user interface and overall experience. This research employed User-Centered Design (UCD), implementing Design Sprint, and Goal-Directed Design (GDD) methodologies to achieve a high value of user satisfaction. The Sprint process is used to create the design prototype efficiently, and the Goal-Directed method is performed to create specific, measurable goals. Our evaluation shows that an iterative design process based on user feedback results in higher user satisfaction. The analysis indicated a significant increase of approximately 25% in both effectiveness and efficiency among users when using the application. Initially, the first iteration reveals a low value of approximately 59% in overall effectiveness and efficiency. However, a substantial boost to around 83.5% was seen after a redesign. This improvement proved that User-Centered Design methodologies highly influence user perception and satisfaction. © 2024 IEEE.

关键词： design sprint goal-directed design skincare UI/UX user-centered design

来源：评论

学校读者我要写书评

暂无评论

Indonesian Tweet Emotion Detection Using IndoBERT 9

Indonesian Tweet Emotion Detection Using IndoBERT

引用

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Masaling, Nikita Ananda Putri Siswanto, Ricky Reynardo Girsang, Abba Suganda Bina Nusantara University BINUS Graduate Program - Master of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350390025

This study investigates emotion detection on tweets written in the Indonesian language, using the IndoBERT model that has been trained for 50 epochs. The examination conducted in this study provides a comprehensive analysis of the performance dynamics shown by the model. The assessment results provide an accuracy score of 0.74, indicating the model's proficiency in identifying sentiments within the collection of tweets. By using assessment criteria such as accuracy, recall, and the F1-score, the model's comprehensive analytical ability is validated, emphasizing its effectiveness in deciphering various emotional circumstances. The accuracy, recall, and F1-score of the model, with average values of 0.79, 0.78, and 0.78 respectively, combined demonstrate its resilience in accurately capturing subtle emotional subtleties. This study enhances our comprehension of language model applications in the field of emotional intelligence by exploring the complexities of sentiment analysis. It emphasizes the practical value of these applications in real-world scenarios. © 2024 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

The future of virtual reality: Prospect and problems 9th

The future of virtual reality: Prospect and problems

引用

9th International Conference on computer science and Computational Intelligence, ICCSCI 2024

作者： Lie, Andrew Alfonso Buntoro, Owen Tamashi Purwanto, Eko Setyo Ario, Muhamad Keenan Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia Mobile Application & Technology Program Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia

Virtual Reality (VR) is a technology that allows users to interact with a simulated environment created by virtual reality capabilities through the intermediary of computers, simulator tools, and others, providing new experiences and can feel experiences that are not possible in the real world. This research aims to identify and analyze what challenges may occur such as technical, social, and ethical barriers in the development and application of VR technology. Using a mixed methods approach, we collected quantitative data from Kaggle's dataset and conducted in-depth interviews with active VR users. Our findings show that although VR has significant potential in various fields, such as education, healthcare, and entertainment, challenges such as motion sickness, high costs, and limited and engaging content hinder the widespread adoption of VR technology. The study also highlights the importance of improving technologies that need to be implemented such as visual capabilities, motion responsiveness, and audio quality to enhance the user experience when using VR. By overcoming these barriers, VR can maximize its positive impact and become a revolutionary tool in various sectors. © 2024 The Authors.

关键词： Virtual environments

来源：评论

学校读者我要写书评

暂无评论

Classification and Comprehension of Software Requirements Using Ensemble Learning

引用

computers, Materials & Continua 2024年第8期80卷 2839-2855页

作者： Jalil Abbas Arshad Ahmad Syed Muqsit Shaheed Rubia Fatima Sajid Shah Mohammad Elaffendi Gauhar Ali School of Computer Science and Technology Anhui UniversityHefei230039China School of Computing Sciences Pak Austria FachhochschuleInstitute of Applied Sciences and TechnologyHaripur22620Pakistan Department of Computer Science and IT University of LahoreLahore55150Pakistan Department of Computer Science Emerson UniversityPunjabMultan60000Pakistan EIAS(Emerging Intelligent Autonomous Systems)Data Science Lab Prince Sultan UniversityRiyad12435Saudi Arabia

The software development process mostly depends on accurately identifying both essential and optional ***,user needs are typically expressed in free-form language,requiring significant time and human resources to translate these into clear functional and non-functional *** address this challenge,various machine learning(ML)methods have been explored to automate the understanding of these requirements,aiming to reduce time and human ***,existing techniques often struggle with complex instructions and large-scale *** our study,we introduce an innovative approach known as the Functional and Non-functional Requirements Classifier(FNRC).By combining the traditional random forest algorithm with the Accuracy Sliding Window(ASW)technique,we develop optimal sub-ensembles that surpass the initial classifier’s accuracy while using fewer *** results demonstrate that our FNRC methodology performs robustly across different datasets,achieving a balanced Precision of 75%on the PROMISE dataset and an impressive Recall of 85%on the CCHIT *** datasets consistently maintain an F-measure around 64%,highlighting FNRC’s ability to effectively balance precision and recall in diverse *** findings contribute to more accurate and efficient software development processes,increasing the probability of achieving successful project outcomes.

关键词： Ensemble learning machine learning non-functional requirements requirement engineering accuracy sliding window

来源：评论

学校读者我要写书评

暂无评论

Enhancing Waste Classification with YOLOv8 Models for Efficient and Accurate Sorting 9th

Enhancing Waste Classification with YOLOv8 Models for Effici...

引用

9th International Conference on computer science and Computational Intelligence, ICCSCI 2024

作者： Vieri, Deverel Susanto, Rendy Purwanto, Eko Setyo Ario, Muhamad Keenan Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia Mobile Application & Technology Program Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia

As of 2023, Indonesia ranks as the second largest waste-producing country globally. The waste produced is not segregated properly, leading to the vast amount of waste piling up in local landfills. Traditional methods, such as manual sorting, have been widely used to segregate waste but suffer from inefficiencies and inaccuracies. In contrast, deep learning models offer an alternative solution for waste classification, overcoming the limitations of traditional methods. A deep learning approach using YOLOv8 was proposed to classify waste into six distinct categories. Three different YOLOv8 variants: nano, small, and medium, are trained after the dataset has been augmented into 3,500 labeled images. The results indicate that these models were able to achieve high accuracy in classifying images, with the nano variant having the least training time and an accuracy of approximately 89%. © 2024 The Authors.

关键词： Contrastive Learning

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：