检索结果-内蒙古大学图书馆

Decomposition and Reorganization of Phonetic Information for Speaker Embedding Learning

IEEE/ACM Transactions on Audio Speech and Language Processing 2023年 31卷 1745-1757页

作者： Hong, Qian-Bei Wu, Chung-Hsien Wang, Hsin-Min National Cheng Kung University and Academia Sinica Graduate Program of Multimedia Systems and Intelligent Computing Tainan701 Taiwan Academia Sinica Taipei115 Taiwan National Cheng Kung University Department of Computer Science and Information Engineering Tainan701401 Taiwan Academia Sinica Institute of Information Science Taipei115 Taiwan

Speech content is closely related to the stability of speaker embeddings in speaker verification tasks. In this paper, we propose a novel architecture based on self-constraint learning (SCL) and reconstruction task (RT) to remove the influence of phonetic information on speaker embedding generation. First, SCL is used to reduce the divergence of frame-level features, which can avoid ambiguity between the resulting embeddings of the two utterances being compared. Second, RT is used to further remove phonetic information in frame-level layers, focusing on speaker-discriminative feature transformation. In our experiments, the speaker embedding models were trained on the VoxCeleb2 dataset and evaluated on the VoxCeleb1, Librispeech, SITW and VoxMovies datasets. Experimental results on VoxCeleb1 show that the proposed DROP-TDNN system reduced the EER by 7.5%, compared to the state-of-the-art ECAPA-TDNN system. Furthermore, the proposed DROP-TDNN system also outperformed the ECAPA-TDNN system in the experiments on SITW, Librispeech and VoxMovies under cross-dataset conditions. In the experiments on SITW, the proposed system reduced the EER by 3.4% compared to the ECAPA-TDNN system. In the experiments on Librispeech, the proposed system demonstrated the advantage of removing phonetic information under the clean speech condition, with a significant reduction of 25.5% in EER compared to the ECAPA-TDNN system. In the experiments on VoxMovies, the proposed system reduced the EER by up to 7.9% compared to the ECAPA-TDNN system under different pronunciation and background conditions. © 2014 IEEE.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Monitoring Applications for Vehicle based on Internet of Things (IoT) using the MQTT Protocol

引用

Procedia computer science 2023年 227卷 73-82页

作者： Nurudin Arief Agus Sukmandhani Muhammad Zarlis Computer Science Department BinusOnline Learning Bina Nusantara University Jakarta Indonesia 11480 Information Systems Management Department BINUS Graduate Program - Master of Information Systems Management Bina Nusantara University Jakarta Indonesia 11480

Monitoring of vehicle conditions is needed as one of the securities in driving because one of the accident rates is negligence in checking the condition of the vehicle. This study aims to help reduce the level of accidents and increase human awareness for vehicle conditions, especially minibus type vehicles. The system works to provide information on the condition of vehicles sourced from the Engine Control Unit (ECU). ECU data cars are obtained by SNIP traffic broadcast message and ask for data directly at the ECU. The car can provide vehicle information using a Canbus communication protocol through the OBD-II socket. Vehicle information can be seen by users through the website in real time. Based on the evaluation from the user side, this system provides ease of monitoring vehicle conditions because it can be accessed via the internet using various kinds of devices.

关键词： OBD-II Blackbox Sstem Logging ECU

来源：评论

学校读者我要写书评

暂无评论

EUDRL: Explainable Uncertainty-Based Deep Reinforcement Learning for Portfolio Management 15

EUDRL: Explainable Uncertainty-Based Deep Reinforcement Lear...

引用

15th International Conference on Computing Communication and Networking Technologies, ICCCNT 2024

作者： Lina, Jia Banda, Shahab S. Paib, Hao-Ting Rawal, Bharat S. International Graduate School of Artificial Intelligence Yunlin County Douliu City Taiwan Bachelor Program of Big Data Applications National Pingtung University Pingtung Taiwan Department of Computer Science and Digital Technologies University GgramblingLA United States

ISBN: (纸本)9798350370249

Supervised deep learning (SDL) has shown remarkable success in various financial applications, such as stock prediction and fraud detection. However, SDL's reliance on class labels renders it unsuitable for portfolio management (PM) tasks, where such labels are often unavailable. To address this limitation, we propose a novel two-level architecture based on deep reinforcement learning (DRL) for PM, which does not require class labels. Our approach comprises several local agents that provide trading decisions and uncertainty assessments for individual stocks, and a global agent that makes portfolio management decisions based on the outputs of the local agents. Additionally, we incorporate the concept of explainable AI (XAI) into our framework using the SHAP (Shapley additive explanations) method, enhancing the transparency and interpretability of the global agent's decisions. Our experimental results demonstrate that the proposed architecture consistently yields profitable outcomes in the market. ©2024 IEEE.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

The Effectivity of Color For Chroma-Key Techniques

引用

Procedia computer science 2021年 179卷 281-288页

作者： Carolus Raditya Muhammad Rizky Sergio Mayranio Benfano Soewito Computer Science Department BINUS Graduate Program-Master of Computer Science Bina Nusantara University Jakarta Indonesia 11480

In this age where almost all facets of life are full of digital objects, because of technology is advancing more and more rapidly, included the fields of economy, the social field, the industrial field, labour, and the entertainment field especially the film industry. In film industry, the backgrounds that were once impossible such as foreign planets and the deep expanses of space are now possible, this is due to the technique known as Chroma Key as well as the usage of green screens and blue screens. In this research we wish to find out if only green screens and blue screens are the only viable colours to use or is there any other colours available to use to Chroma *** use or are other colours available to use to Chroma Key. The methods that we used to conduct research on this topic are case studies. We conduct a case study to see the level of efficiency and fatigue after using various colours as a background to the Chroma Key process. Experiments: We ourselves will try to conduct the Chroma Key process while using various colours as the background and making sure that other variables such as our lighting and spacing are constant on every case we conducted, to measure the efficiency and effectivity of the colours in the Chroma Key process.

关键词： colour chroma-key effectivity of colour background

来源：评论

学校读者我要写书评

暂无评论

Implementation of expert systems in potassium deficiency in cocoa plants using forward chaining method

引用

Procedia computer science 2023年 216卷 136-143页

作者： Muhammad Titan Hafizal Dian Pratama Putra Herry Wirianata Nanda Satya Nugraha Teddy Suparyanto Alam Ahmad Hidayat Bens Pardamean Faculty of Forestry STIPER Institute of Agriculture Yogyakarta 55282 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta 11480 Indonesia Computer Science Department BINUS Graduate Program Bina Nusantara University Jakarta 11480 Indonesia

As one of the largest exporters in the world, cocoa (Theobroma cacao L.) production in Indonesia provides an important contribution to the plantation sector that can, directly and indirectly, attribute to the national economic development. However, recent data shows a decline in cocoa productivity in Indonesia. This is arguably caused by various factors, including poor agricultural management and practices in cocoa plantations, that may lead to lower fertility in the soil and an increased risk of diseases and pests. Potassium deficiency is a major contributing factor to the low soil fertility that affects cocoa yields. Therefore, in this work, we implement an application with an expert system utilizing the forward chaining method to detect potassium deficiency in cocoa plants and then give a fertilization-based recommendation based on the plants’ condition. The system employs a set of rules to identify symptoms related to the deficiency on the sample photo of a cocoa leaf according to the channels of red, green, and blue of the image. The sample images of cocoa leaves are submitted to the application with an easy-to-use interface that can show the scanning result and proceed to display the suggested quantity of fertilizers to prevent potassium deficiency. Implementing the system can contribute constructive impacts to improve current practices in the overall cropping system of cocoa plants.

关键词： agriculture cocoa expert system forward chaining recommendation system

来源：评论

学校读者我要写书评

暂无评论

Performance of Information Technology Infrastructure Prediction using Machine Learning

引用

Procedia computer science 2021年 179卷 515-523页

作者： Ignatius Rahardjo Heruwidagdo Suharjito Novita Hanafiah Yanto Setiawan Computer Science Department Binus Graduate Program-Master in Computer Science Bina Nusantara University Indonesia Computer Science Department BINUS Online Learning Bina Nusantara University Jakarta Indonesia 11480

Resource management is always an important issue related to good governance decision making. One of the common problem faced in managing IT Infrastructure is about allocating server resources to improve the performance. In this study we use a machine learning approach to make predictions about the performance of information technology infrastructure. The experiment took data from several servers in a company to be tested. The performance measure of resources used in this study are CPU Performance, Disk performance, Memory capacity, and Network performance. Several algorithms and machine learning methods are tested, such as Linear Regression, kNN, SVR, Decision Tree and Random Forest, to find the best model fit for the servers. The comparison result shows that Linear regression and kNN perform well in predicting the network performance in those three servers.

关键词： log performance machine learning kNN IT infrastructure performance prediction

来源：评论

学校读者我要写书评

暂无评论

Lockdown strategy worth lives: The SEIRD modelling in COVID-19 outbreak in Indonesia 2

Lockdown strategy worth lives: The SEIRD modelling in COVID-...

引用

2nd International Conference on Biospheric Harmony Advanced Research, ICOBAR 2020

作者： Nurlaila, I. Hidayat, A.A. Pardamean, B. Information Systems Department BINUS Online Learning Bina Nusantara University Jakarta11480 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia Computer Science Department BINUS Graduate Program-Master of Computer Science Bina Nusantara University Jakarta11480 Indonesia

Lockdowns, despite their conflicting restrictions and consequences they might offer when enforced as a national strategy, are deemed to be suggestive for a prompt conquer to the Coronavirus Disease-19 (COVID-19) outbreak. There have been some success stories such as in China, South Korea and our ASEAN member fellow, Vietnam, which exhibited extremely fallen numbers of COVID-19 cases post the enforcement of lockdown. Indonesia, however, remains in the crux of dispute whether or not the lockdown is opted to force COVID-19 transmission down under control. We, in this respect, employ the most popular model which has been broadly applied in the field of epidemiology, referred to as SEIRD (Susceptible, Exposed, Infectious, Recovered, and Death), the extension form of an age-structured SEIR, where the Death (D) is included to provide more factual situation. We modify the I (Infectious) fraction as symp-tomatic (Is) and asymptomatic (Ia) infectives. There are three lockdown scenarios simulated in our modified SEIRD with the starting date are 26 April 2020 when the immediate lockdown was enacted-and a-week and two-week-gaps, respectively. We figured out that by stipulating the lockdown without delay (26 April 2020), the new cases could be kept below 10,000. A week delay escalates case number to 5,000 and delay for one more week results in beyond 20,000 new cases. Furthermore, when the intervention is delayed following the delayed lockdown, normalization would demand a longer period within which dealing with more critical and dying patients is unavoidable and should be more anticipated. © Published under licence by IOP Publishing Ltd.

关键词： Coronavirus

来源：评论

学校读者我要写书评

暂无评论

The Deep Learning Approach For American Sign Language Detection

The Deep Learning Approach For American Sign Language Detect...

引用

Creative Communication and Innovative Technology (ICCIT), IEEE

作者： Gusti Pangestu Yaya Heryadi Lukas Wayan Suparta Yulyani Arifin Computer Science Department BINUS Graduate Program Computer Science. Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program Computer Science Bina Nusantara University Jakarta Indonesia Cognitive Engineering Research (CERG) Faculty of Engineering Universitas Katolik Indonesia Atma Jaya Jakarta Indonesia Department of Electrical Engineering Faculty of Industrial Technology Institut Teknologi Nasional Yogyakarta Yogyakarta Indonesia

Sign language has importance rule to deal with communication process especially with impairments hearing people. Sign language detection also attract lot of researchers to join the challenge of research to detect and recognize the sign language in the field of computer science. Hence, there is still no any standard approach and method to recognize the meaning in every pose of sign language. This research proposed a mechanism to detect Alphabet American Sign Language by utilizing Convolutional Neural Network (CNN) process. The CNN approach was chosen based on the ability and capability to recognize image. In this research, MNIST dataset is used for traning and testing process. The proposed CNN architecture produced 97% of accuracy that outperform the previous research using the same dataset which made this architecture promising.

关键词： Deep learning computer science Image recognition Gesture recognition computer architecture Auditory system Assistive technologies

来源：评论

学校读者我要写书评

暂无评论

Long short-term memory (LSTM) model-based reinforcement learning for nonlinear mass spring damper system control

引用

Procedia computer science 2023年 216卷 213-220页

作者： Santo Wijaya Yaya Heryadi Yulyani Arifin Wayan Suparta Lukas Computer Science Department BINUS Graduate Program – Doctor of Computer Science Bina Nusantara University Jakarta 11480 Indonesia Department of Electrical Engineering Faculty of Industrial Technology Institut Teknologi Nasional Yogyakarta Yogyakarta 55281 Indonesia Cognitive Engineering Research Group (CERG) Universitas Katolik Indonesia Atma Jaya Jakarta 12930 Indonesia

The Neural Networks (NN) model which is incorporated in the control system design has been studied, and the results show better performance than the mathematical model approach. However, some studies consider that only offline NN model learning and does not use the online NN model learning directly on the control system. As a result, the controller's performance decreases due to changes in the system environment from time to time. The Reinforcement Learning (RL) method has been investigated intensively, especially Model-based RL (Mb-RL) to predict system dynamics. It has been investigated and performs well in making the system more robust to environmental changes by enabling online learning. This paper proposes online learning of local dynamics using the Mb-RL method by utilizing Long Short-Term Memory (LSTM) model. We consider Model Predictive Control (MPC) scheme as an agent of the Mb-RL method to control the regulatory trajectory objectives with a random shooting policy to search for the minimum objective function. A nonlinear Mass Spring Damper (NMSD) system with parameter-varying linear inertia is used to demonstrate the effectiveness of the proposed method. The simulation results show that the system can effectively control high-oscillating nonlinear systems with good performance.

关键词： Model-based Reinforcement Learning Random-Shooting Policy LSTM MPC Nonlinear System

来源：评论

学校读者我要写书评

暂无评论

Privacy-Enhanced Data Deduplication Computational Intelligence Technique for Secure Healthcare Applications

引用

computers, Materials & Continua 2022年第2期70卷 4169-4184页

作者： Jinsu Kim Sungwook Ryu Namje Park Department of Convergence Information Security Graduate SchoolJeju National University63294Korea Master’s Program in Future Strategy Korea Advanced Institute of Science and Technology34141Korea Department of Computer Education Teachers CollegeJeju National University63294Korea

A significant number of cloud storage environments are already implementing deduplication *** to the nature of the cloud environment,a storage server capable of accommodating large-capacity storage is *** storage capacity increases,additional storage solutions are *** leveraging deduplication,you can fundamentally solve the cost ***,deduplication poses privacy concerns due to the structure *** this paper,we point out the privacy infringement problemand propose a new deduplication technique to solve *** the proposed technique,since the user’s map structure and files are not stored on the server,the file uploader list cannot be obtained through the server’s meta-information analysis,so the user’s privacy is *** addition,the personal identification number(PIN)can be used to solve the file ownership problemand provides advantages such as safety against insider breaches and sniffing *** proposed mechanism required an additional time of approximately 100 ms to add a IDRef to distinguish user-file during typical deduplication,and for smaller file sizes,the time required for additional operations is similar to the operation time,but relatively less time as the file’s capacity grows.

关键词： Computational intelligence cloud multimedia data deduplication

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：