检索结果-内蒙古大学图书馆

AVTENet: A Human-Cognition-Inspired Audio-Visual Transformer-Based Ensemble Network for Video Deepfake Detection

IEEE Transactions on Cognitive and Developmental Systems 2025年

作者： Hashmi, Ammarah Shahzad, Sahibzada Adil Lin, Chia-Wen Tsao, Yu Wang, Hsin-Min Taiwan International Graduate Program Academia Sinica Social Networks and Human-Centered Computing Program Taipei11529 Taiwan National Tsing Hua University Institute of Information Systems and Applications Hsinchu30013 Taiwan National Chengchi University Department of Computer Science Taipei11605 Taiwan National Tsing Hua University Department of Electrical Engineering Institute of Communications Engineering Hsinchu300044 Taiwan Academia Sinica Research Center for Information Technology Innovation Taipei11529 Taiwan

The recent proliferation of hyper-realistic deepfake videos has drawn attention to the threat of audio and visual forgeries. Most previous studies on detecting artificial intelligence-generated fake videos only utilize visual modality or audio modality. While some methods exploit audio and visual modalities to detect forged videos, they have not been comprehensively evaluated on multimodal datasets of deepfake videos involving acoustic and visual manipulations, and are mostly based on convolutional neural networks with low detection accuracy. Considering that human cognition instinctively integrates multisensory information including audio and visual cues to perceive and interpret content and the success of transformer in various fields, this study introduces the audio-visual transformer-based ensemble network (AVTENet). This innovative framework tackles the complexities of deepfake technology by integrating both acoustic and visual manipulations to enhance the accuracy of video forgery detection. Specifically, the proposed model integrates several purely transformer-based variants that capture video, audio, and audio-visual salient cues to reach a consensus in prediction. For evaluation, we use the recently released benchmark multimodal audio-video FakeAVCeleb dataset. For a detailed analysis, we evaluate AVTENet, its variants, and several existing methods on multiple test sets of the FakeAVCeleb dataset. Experimental results show that the proposed model outperforms all existing methods and achieves state-of-the-art performance on Testset-I and Testset-II of the FakeAVCeleb dataset. We also compare AVTENet against humans in detecting video forgery. The results show that AVTENet significantly outperforms humans. © 2016 IEEE.

关键词： audio-visual AVTENet Deepfake detection

来源：评论

学校读者我要写书评

暂无评论

Multi-Criteria based Decision Model for Evaluation 6

Multi-Criteria based Decision Model for Evaluation

引用

6th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2023

作者： Utama, Ditdit Nugeraha Bin Widayat, Muhamad Akbar Bina Nusantara University BINUS Graduate Program Master of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350358346

Addressing multiple criteria and parameter issues in computer modelling presents a significant challenge. Several factors including data types, parameter behaviours, and purposes, must be taken into account to enhance computer modelling capability;particularly in evaluation cases. Through the utilization of a multi-criteria and method approach, a decision model was effectively developed to assess a case of environmental sustainability level of a building. One method operated in the study is the curve method for handling membership function form in realizing fuzzy logic. This innovative model demonstrates superior performance. It achieves an impressive accuracy rate of 96%, surpassing the previous model that employed a trapezoidal approach to describe fuzzy membership functions hy 1%. © 2023 IEEE.

关键词： Fuzzy logic

来源：评论

学校读者我要写书评

暂无评论

Business Intelligence Model to Analyze Social Media Information 3

Business Intelligence Model to Analyze Social Media Informat...

引用

3rd International Conference on computer science and Computational Intelligence, ICCSCI 2018

作者： Kurnia, Parama Fadli Suharjito Computer Science Department BINUS Graduate Program-Master of Computer Science Bina Nusantara University Jakarta11480 Indonesia

Social media is a platform to share information that is very liked by everyone nowadays because some of the facilities that make it easier for us to communicate with each other, share documents, chat and even create a community. In addition, we can also analyze the content of social media by using several methods in data mining, so that we can get new the information to support decision making that can bring benefits to individuals and companies. The purpose of this research, to create a business intelligence dashboard to observe the performance of each Topic or channel of news posted to social media accounts such as Facebook and Twitter. Topical performance in social media is the number of Topics in articles posted to social media getting like, share, comment etc. To be able to know the Topic of a news post in social media, used some text classification techniques such as Naive Bayes, SVM and Decision Tree. The comparative results of the algorithms are taken which has the best accuracy of SVM for subsequent implementation in the data warehouse. Meanwhile, the business intelligence dashboard data source will be sourced from the data warehouses that have been made before. © 2018 The Authors. Published by Elsevier Ltd.

关键词： Data warehouses

来源：评论

学校读者我要写书评

暂无评论

Song recommendation system using collaborative filtering methods 3

Song recommendation system using collaborative filtering met...

引用

3rd International Conference on Digital Technology in Education, ICDTE 2019

作者： Girsang, Abba Suganda Wibowo, Antoni Edwin Computer Science Department BINUS Graduate Program - Master of Computer Science Bina Nusantara University Jakarta11480 Indonesia

ISBN: (纸本)9781450372206

Currently, digital online music increase significantly, both in terms of content and users. Increasing the number of digital music content every month conduce a lot of song catalog data and becoming unstructured and making it difficult for users to choose the songs they want to listen to. To make it easier for users to optimize a large number of subscribed music catalogs, a user-centric music recommendation system is needed that allows users to be able to manage catalogs of digital music content according to their needs. This study examines how to implement song recommendation system using collaborative filtering method in digital online music. This study uses 20,000 users, 6,000 songs and 470,000 transactions rating. Through those research, it is discovered that user-based collaborative technique that could make one system for clients will gather those playlist they really want to hear. © 2019 Association for Computing Machinery

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Sales Distribution Activity Tracking System Using Mobile Location Based Services 3

Sales Distribution Activity Tracking System Using Mobile Loc...

引用

3rd International Conference on Information Management and Technology, ICIMTech 2018

作者： Daryanto Suharjito Computer Science Department BINUS Graduate Program-Master of Computer Science Bina Nusantara University Jakarta11480 Indonesia

ISBN: (纸本)9781538658215

The current application sales distribution is still not effective because the officer in the field cannot open the application and still use manual note for distribution sales activities. To overcome this, there is need for development to the mobile application. Sales Distribution Activity Tracking System Using Mobile Location Based Services considered very suitable for developing a web application sales distribution. Applications will be installed on android mobile devices associated with databases and servers equipped with the notification. With this application, field officers can input transactions through its android mobile devices. This research will design and build mobile applications sales distribution on smartphone android based. Applications built using Waterfall process model with development programming language C #, SQLite database and use software Xamarin Studio IDE. These applications can run on smartphone with android version 5.0 to 6.0. The results of this study have designed and built an applications that can be used by the company for better process management of the sales distribution and can find sales reports are up to date. The sales distribution activity tracking system using location-based mobile services makes it easy for sales distributions to report their activities to a system that was previously done manually. Based on the result of system feasibility evaluation, this application can be categorized as feasible for business. © 2018 IEEE.

关键词： Location based services

来源：评论

学校读者我要写书评

暂无评论

Neural Network and Resampling to Handle Imbalance Class on Estimate Software Defect 15

Neural Network and Resampling to Handle Imbalance Class on E...

引用

15th International Congress on Advanced Applied Informatics Winter, IIAI-AAI-Winter 2023

作者： Setiawan, Tri Agus Soeparno, Haryono Gaol, Ford Lumban Arifin, Yulyani Bina Nusantara University Binus Graduate Program - Doctor of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350383829

Rapid progress in the field of artificial intelligence has created opportunities for extensive applications in software development. One area that receives attention is the evaluation of code quality using machine learning techniques. In this investigation, we examined the possible application of machine learning to predict the likelihood of defects in computer code. We employ NASA archival data as case studies. Machine learning models employ neural network algorithms. Our exploration involves partitioning the dataset into training data and test data for performance evaluation. The findings indicate that the Neural Organize technique with resampling yields a high level of accuracy in predicting software defects. Our simulated neural network is capable of identifying intricate patterns in the data and providing precise measurements of the size and intensity of defects. These results have significant implications in the software business, enabling developers to promptly identify possible vulnerabilities and take preventive measures before product release. © 2023 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Deep learning and Threshold Probability for Out of Scope Intent Detection in Task Oriented Chatbot 3

Deep learning and Threshold Probability for Out of Scope Int...

引用

3rd International Conference on Artificial Intelligence and Data sciences, AiDAS 2022

作者： Hasani, Muhammad Fikri Gaol, Ford Lumban Soewito, Benfano Warnars, Harco Leslie Hendric Spits Binus Graduate Program - Doctor of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (数字)9781665491648

ISBN: (纸本)9781665491648

A lot of research on task-oriented chatbots has been carried out to improve research on intent recognition. However, there are not many studies that discuss out-of-scope. This study introduces the use of threshold for classification results. The use of threshold improves recall performance of OOS detection above baseline for BiLSTM but reduces recall inside scope for all models. Directly applying multiclass classification with combined class of inside scope and out of scope also showed higher recall than baseline. © 2022 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Recommendation System Model for Personalized Learning in Higher Education using Content-Based Filtering Method 7

Recommendation System Model for Personalized Learning in Hig...

引用

7th International Conference on Information Management and Technology, ICIMTech 2022

作者： Julianti, M. Ramaddan Heryadi, Yaya Yulianto, Budi Budiharto, Widodo BINUS Graduate Program - Doctor of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (数字)9781665450904

ISBN: (纸本)9781665450904

- An online learning system is a learning approach that typically uses a one fit for all approach in which all student abilities are considered the same so that the provision of learning materials provided is the same for all students. In contrast, a personalization learning system is a teaching and learning approach in which students are positioned as the learning centers and the students are motivated and facilitated to overcome their learning needs based on their learning speed and understanding of the learning process carried out to achieve. In the context of learning, learning materials and learning objectives are the main components that are interrelated in achieving the expected learning outcomes. However, the relationship between learning materials, especially in prerequisite courses, must have similarities so that the learning process can continue in accordance with the learning outcomes to be achieved. This paper presents an experiment results of the proposed the recommendation system modeling for personalization learning model implemented using Content-Based Collaborative Filtering model. In the proposed recommender system model, similarities between the learning materials and the description of the course being are computed as basis for developing recommended learning materials to students. The recommendation system modeling for personalized learning is applied as a solution to achieve learning goals and minimize dropouts due to the inability of students to understand the learning process in online learning systems. The results given in the research are modeling of an adaptive learning personalization recommendation system to provide recommendations for students who do not reach the minimum competency standards. © 2022 IEEE.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Business Intelligence for Product Defect Analysis 6

Business Intelligence for Product Defect Analysis

引用

Joint Conference of the 6th Annual Conference of Industrial and System Engineering 2019, ACISE 2019 and 1st International Conference on Risk Management as an Interdisciplinary Approach 2019, ICRMIA 2019

作者： Girsang, A.S. Isa, S.M. Haris, A.L. Arwan Mandagie, K. Ariana, L.R. Ardinda, V. Computer Science Department BINUS Graduate Program-Master of Computer Science Bina Nusantara University Jakarta11480 Indonesia

The problem of reporting service in *** is the complexity of the query to take the data from the table in oracle. There are many table to be processed for the data, because of that it's hard to show the report quickly. PT XYZ needs fast and reliable report to be shown, especially for control or audit. The suggested solution is to use data warehouse, because it has more structured storage and the data can be analysed by OLAP analysis that has the capability to show the data quickly. Data warehouse development was done through nine step that has been designed by Kimball and Ross. With this solution, PT XYZ has the capability to maintain quality of the product by monitoring the process and the data. © Published under licence by IOP Publishing Ltd.

关键词： Data warehouses

来源：评论

学校读者我要写书评

暂无评论

Service Oriented Fuzzy Decision Model for Health Insurance Claim 5

Service Oriented Fuzzy Decision Model for Health Insurance C...

引用

5th International Conference on Advancement in Data science, E-Learning and Information Systems, ICADEIS 2023

作者： Utama, Ditdit Nugeraha Susanto, Sumiatie Computer Science Department BINUS Graduate Program Master of Computer Science Bina Nusantara University Jakarta11480 Indonesia

ISBN: (纸本)9798350303414

Insurance claim is a fascinating issue to study. A potential loss for the insurance company is major coming from this issue. Thus, many studies performed already to answer such an issue. This study takes aim to develop a computational decision model based on service technology. Four fundamental methods operated in constructing the model and its implementation in service technology. The analytical hierarchical process (AHP) and fuzzy logic methods are two methods benefited to construct the model;where the AHP used to prioritize thirteen parameters considered and the fuzzy logic with its inference capability operated to generate the decision. Object oriented is an analysis and design method to analyze and design the model implanting it in service oriented architecture (SOA). Then, SOA conception functioned to deploy the model in the service architecture. Ultimately, the suggested framework comprising three layers of service-oriented architecture (SOA), namely business process, service interface, and application, has been established, alongside the integration of eight essential services that connect these three applications. The model demonstrates simulation outcomes indicating that 31.47% of claims are categorized as low risk and have been approved, 17.64% of claims are considered moderate risk with currently pending decision status (requiring additional investigation), while 50.89% of claims are classified as high risk with also pending decision status. © 2023 IEEE computer Society. All rights reserved.

关键词： Fuzzy logic

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：