检索结果-内蒙古大学图书馆

Evaluation of phone posterior probabilities for pathology detection in speech data using deep learning models

International Journal of Speech Technology 2025年第1期28卷 99-116页

作者： Farazi, Sahar Shekofteh, Yasser Faculty of Computer Science and Engineering Shahid Beheshti University Tehran Iran

Voice pathology detection (VPD) aims to accurately identify voice impairments by analyzing speech signals. This study proposes models based on deep learning (DL) for binary classification to distinguish between healthy and pathological voices, with a unique focus on the integration of phone posterior probabilities (PPPs as phonetic-based features) alongside Mel-frequency cepstral coefficients (MFCCs as acoustic-based features) for input models. By incorporating PPPs as supplementary information, we investigate the model’s performance across spontaneous, sustained vowel, and read speech datasets, addressing the gap in comparing these speech types for VPD. Our results highlight that PPPs significantly enhance classification accuracy, particularly for read speech and spontaneous speech data types. Using the AVFAD database, we show that the proposed CNN-based model achieves its highest performance on spontaneous speech, with an accuracy of approximately 87% on test data and 93% on validation data. This study emphasizes the impact of PPPs as phonetic-based features in VPD tasks and clarifies which types of speech benefit most from their inclusion, paving the way for more refined models in this field. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： AVFAD dataset Convolutional neural network Deep learning Phone posterior probabilities Spontaneous speech Voice pathology detection

来源：评论

学校读者我要写书评

暂无评论

APEDM: a new voice casting system using acoustic–phonetic encoder-decoder mapping

引用

Multimedia Tools and Applications 2024年 1-25页

作者： Alipour Esgandani, Sogol Shekofteh, Yasser Moradi, Ashkan Faculty of Computer Science and Engineering Shahid Beheshti University Tehran Iran

Voice casting is one of the most crucial aspects of dubbing and localization, which consists of adapting audio-visual content from one language and culture to another. Dubbing is widely used in various industries, such as entertainment, education, marketing, and gaming, to reach global audiences and increase customer satisfaction. Choosing the right voice actor is not a simple task that anyone can do and requires consideration of many aspects of the source (original language) and target material (in the dubbed language), such as language characteristics, acoustic characteristics of the characters, age, emotions, and gender. An automated system can be a powerful tool alongside directors/managers to improve the selection process. These systems rely on the acoustic characteristics of the speakers and are powered by machine learning methods to achieve reliable results. This paper intends to contribute by introducing a novel method for nonlinear mapping voices between languages using encoder-decoder neural networks and phonetic-acoustic features. It details the architecture of the proposed automatic voice casting (AVC) system and reports experimental results comparing it with other baseline frameworks. In this study, the proposed AVC system is for English to Farsi voice casting in closed map situations. The experiments are conducted using the English to Persian voice-casting (E2PCast) dataset. The baseline system based on siamese neural networks (SNN) and x-vector input had a 4-best character accuracy of 99.25% and an accuracy of 96.32% on the test data. Our proposed system, acoustic–phonetic encoder-decoder mapping (APEDM), with an x-vector and phonetic vector as its inputs while applying SNN for similarity scoring showed the best results. This system achieved 98.98% accuracy and 4-best character accuracy of 100% in the test data. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Customer satisfaction

来源：评论

学校读者我要写书评

暂无评论

Voice pathology detection on spontaneous speech data using deep learning models

引用

International Journal of Speech Technology 2024年第3期27卷 739-751页

作者： Farazi, Sahar Shekofteh, Yasser Faculty of Computer Science and Engineering Shahid Beheshti University Tehran Iran

Speech problems are a common issue that affects people everywhere and can affect the quality of their lives. The human speech production system involves various components. Dysfunction of any of these components can disrupt normal speech production, giving rise to speech diseases like laryngopharyngeal reflux, vocal cord paralysis, and vocal fold nodules. Early diagnosis of these disorders is very important for the patient's health. Many studies in automatic diagnosis of voice pathology have used sustained vowel sounds and read-speech as the primary source of speech data. However, it is crucial to recognize the unique value of spontaneous-speech. In addition to inheriting the characteristics of read speech, spontaneous-speech offers a more authentic glimpse into individuals' speech behavior. It captures not only linguistic features, but also subtle nuances of human emotions, such as fatigue and excitement, which may cause speech impairments, and shows their patterns in the speech signal better than in the read-speech data. Therefore, we aim to explore spontaneous speech in voice pathology detection to determine if it can help us better understand speech problems. In this research, we examine different deep learning (DL) models trained on two main features (MFCC and Mel spectrograms) for binary classification of healthy speech versus pathological speech, with a specific focus on the spontaneous speech. Extensive experimentation reveals the superiority of our proposed convolutional neural network (CNN) model trained on MFCC features. Notably, the CNN model achieves the highest accuracy, approximately 85% for test data and 92% for evaluation data. These results emphasize the potential of DL approaches in the accurate diagnosis of speech disorders through the analysis of the spontaneous-speech, offering promise for early detection and improved patient care. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Doctor Code: A machine learning-based approach to program repair

引用

Scientia Iranica 2024年第2期31卷 83-102页

作者： Moosavi, Sh Vahidi-Asl, M. Haghighi, H. Rezaalipour, M. Faculty of Computer Science and Engineering Shahid Beheshti University Tehran Iran

In t he last decade, several Automated Program Repair (APR) techniques have been proposed. Most of these techniques produce candidate patches using repair operators that apply changes to buggy programs. Therefore, to repair more bugs, a more extensive set of repair operators is required. However, more repair operators lead to longer repair times and more overfitted patches. To address the above-mentioned issues, we present Doctor Code, a new APR technique that chooses repair operators by systematically learning from the features of the most common bugs in different programs based on machine learning. We compare our technique against Mutation repair by the Siemens suite. The experiment results indicate that our technique can fix II bugs while the baseline only repairs 22. In addition. Doctor Code can produce patches that do not exist in the search space of SPR. Prophet, and SemFix. We also tested Doctor Code utilizing three buggy versions of a program called Space to indicate its capability of repairing large-sized programs. Also, we compare Doctor Code against 7 state-of-the-art APR tools, like Elixir, using the Defects4j dataset. The experiment results indicate that our technique outperforms the other tools in terms of the number of fixed bugs and overfitted patches. © 2024 Sharif University of Technology.

关键词： Software testing

来源：评论

学校读者我要写书评

暂无评论

Robust security risk estimation for android apps using nearest neighbor approach and hamming distance

引用

Soft Computing 2025年第2期29卷 593-611页

作者： Deypir, Mahmood Zoughi, Toktam Faculty of Computer Engineering Shahid Sattari Aeronautical University of Science and Technology Tehran Iran Tehran Iran

Nowadays, Android-based devices such as smart phones, tablets, smart watches, and virtual reality headsets have found increasing use in our daily lives. Along with the development of various applications for these devices, new malicious apps are released by intruders, which are more difficult to identify and deal with because they exploit more sophisticated techniques. Although methods have been provided to calculate the security risk and identify malicious apps in Android operating system, but with the expansion of the level and depth of the threats, the need for more effective methods in this context is still required. In this paper, we have devised a new algorithm to calculate the security risk of Android apps, which can be used to identify malicious apps from benign ones. In this algorithm, to estimate the security risk of an unknown input app, its nearest neighbors to malicious apps and its nearest neighbors to normal apps are computed separately using Hamming distance. Then, the security risk of the input app can be computed using a simple formulation. After implementing this algorithm, its parameter for the number of neighbors using various real datasets is adjusted. The extensive experiments conducted on these data show the superiority of the proposed method over the previously proposed ones in terms of detection rate. Our additional experimentations show the robustness of the proposed algorithm in adversarial situations. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Android malware

来源：评论

学校读者我要写书评

暂无评论

Optimized COVID-19 detection using sparse deep learning models from multimodal imaging data

引用

Multimedia Tools and Applications 2025年第13期84卷 11451-11474页

作者： Moradi, MohammadMahdi Hassanzadeh, Alireza Haghanifar, Arman Ko, Seok Bum Faculty of Computer Science and Software Engineering Concordia University MontrealQC Canada Faculty of Electrical Engineering Shahid Beheshti University Tehran Tehran Iran Department of Electrical & Computer Engineering University of Saskatchewan SaskatoonSK Canada

This paper explores the global spread of the COVID-19 virus since 2019, impacting 219 countries worldwide. Despite the absence of a definitive cure, the utilization of artificial intelligence (AI) methods for disease diagnosis has demonstrated commendable effectiveness in promptly diagnosing patients and curbing infection transmission. The study introduces a deep learning-based model tailored for COVID-19 detection, leveraging three prevalent medical imaging modalities: computed tomography (CT), chest X-ray (CXR), and Ultrasound. Various deep Transfer Learning Convolutional Neural Network-based (CNN) models have undergone assessment for each imaging modality. For each imaging modality, this study has selected the two most accurate models based on evaluation metrics such as accuracy and loss. Additionally, efforts have been made to prune unnecessary weights from these models to obtain more efficient and sparse models. By fusing these pruned models, enhanced performance has been achieved. The models have undergone rigorous training and testing using publicly available real-world medical datasets, focusing on classifying these datasets into three distinct categories: Normal, COVID-19 Pneumonia, and non-COVID-19 Pneumonia. The primary objective is to develop an optimized and swift model through strategies like Transfer Learning, Ensemble Learning, and reducing network complexity, making it easier for storage and transfer. The results of the trained network on test data exhibit promising outcomes. The accuracy of these models on the CT scan, X-ray, and ultrasound datasets stands at 99.4%, 98.9%, and 99.3%, respectively. Moreover, these models’ sizes have been substantially reduced and optimized by 51.93%, 38.00%, and 69.07%, respectively. This study proposes a computer-aided-coronavirus-detection system based on three standard medical imaging techniques. The intention is to assist radiologists in accurately and swiftly diagnosing the disease, especially during the screen

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

A Deep Reinforcement Learning Approach Combining Technical and Fundamental Analyses with a Large Language Model for Stock Trading 14

A Deep Reinforcement Learning Approach Combining Technical a...

引用

14th International Conference on computer and Knowledge engineering, ICCKE 2024

作者： Veisi, Mahan Berangi, Sadra Khojasteh, Mahdi Shahbazi Salimi-Badr, Armin Faculty of Computer Science and Engineering Shahid Beheshti University Tehran Iran

ISBN: (纸本)9798331511272

Stock trading strategies are essential for successful investment, yet developing a profitable approach is challenging due to the stock market’s complex and dynamic nature. This paper introduces a Deep Reinforcement Learning (DRL) framework for automated stock trading that integrates technical and fundamental analyses with a large language model. We model the trading environment as a Partially Observable Markov Decision Process (POMDP) and propose a hybrid architecture that combines Long Short-Term Memory (LSTM) with Proximal Policy Optimization (PPO) to capture intricate temporal patterns in stock data and make informed trading decisions. Our model incorporates market indicators alongside financial news headlines, processed through the FinBERT language model, to create a rich state representation. Additionally, we integrate a drawdown penalty into the reward function to further improve portfolio stability. Evaluations on a dataset of 30 U.S. stocks demonstrate that our model outperforms benchmarks in cumulative return, maximum earning rate, and Sharpe ratio, indicating that the hybrid approach yields more resilient and profitable trading strategies than existing methods. ©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

Enhancing Decisions of Goalkeeper and Kicker Players in the RoboCup 2D Simulation League through Behavioral Cloning 20

Enhancing Decisions of Goalkeeper and Kicker Players in the ...

引用

20th CSI International Symposium on Artificial Intelligence and Signal Processing, AISP 2024

作者： Nasiri, Mohammad Hesam Zonouzi, Seyed Hassan Majid Salimi-Badr, Armin Shahid Beheshti University Faculty of Computer Science and Engineering Tehran Iran

ISBN: (纸本)9798350383942

RoboCup is recognized as a landmark project in the robotics and artificial intelligence industry, providing a crucial platform for the advancement of research and innovations in autonomous and multi-agent systems. This study delves into the use of behavioral cloning in the RoboCup 2D Simulation League to imitate the best decisions and improve the performance of the goalkeeper along with the kicker players. To improve the performance of our soccer team in the league, we have developed a new approach that combines behavioral cloning and game log parsing. To achieve this, we played extensively against the champion team Helios and analyzed the actions and decisions of their goalkeeper and kicker players. This data was used to create a large dataset that our team could use to improve their imitation skills. Our innovative log generator tool and game logs parser made the labeling process easier and will also be useful for future research in the league. Novel features are extracted to represent the state of the game considering the limitations of each player including their head direction along with their field of view. Next, a neural structure is trained to imitate the behavior of kicker players and the goalkeeper. Moreover, the time limit for decision-making in each turn is considered. This study showed that using behavioral cloning for imitation learning noticeably improved our team's actions and decisions due to the enhancement of the number of accurate passes, on-target shoots, and saves by the goalkeeper. © 2024 IEEE.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Cluster Sampling: A Cluster-Driven Sampling Strategy for Deep Metric Learning 14

Cluster Sampling: A Cluster-Driven Sampling Strategy for Dee...

引用

14th International Conference on computer and Knowledge engineering, ICCKE 2024

作者： Rafiee, Hamideh Abin, Ahmad Ali Majd, Seyed Soroush Faculty of Computer Science and Engineering Shahid Beheshti University Tehran Iran

ISBN: (纸本)9798331511272

Deep metric learning has gained significant attention recently due to its promising performance in image retrieval, face recognition, and clustering tasks. Deep metric learning algorithms map the original data from the initial feature space to a new embedded space by learning a mapping function, where the discriminative power of the samples is increased. Sampling strategies are a critical aspect of deep metric learning algorithms, as they determine the choice of training samples that shape the network’s learning process. This research introduces a novel sampling approach designed to optimize this selection for improved model performance. Given the high dimensionality of embedded spaces, choosing a sampling strategy is critical for enhancing the quality of learned representations and improving model performance. However, existing sampling methods often face significant challenges, such as slow convergence and the tendency to get trapped in suboptimal local minima. To address these challenges, this research proposes a novel sampling approach that leverages the clustering of sample representations within training batches to identify and select the most informative samples for effective training. This approach ensures a more dynamic learning process by clustering data based on shared features and selecting samples both within clusters and between related clusters. The selected samples enable the model to better learn the underlying data structure, thereby enhancing the discriminative power and robustness of the learned metrics. The effectiveness of the proposed approach is validated through empirical evaluations on the cars196 dataset, where it outperforms existing sampling methods. ©2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Implementing a Serious Traffic Game to Improve Children's Road Awareness Skills

Implementing a Serious Traffic Game to Improve Children's Ro...

引用

2023 International Conference on Electrical and Information Technology, IEIT 2023

作者： Nahvi, Azadeh Mosharraf, Maedeh Shahid Beheshti University Computer Engineering Tehran Iran Shahid Beheshti University Faculty of Computer Science and Engineering Tehran Iran

ISBN: (纸本)9798350327298

To ensure children's safety in streets, it is crucial to teach them the basic rules of traffic. Given the risks associated with training in the real world and the ineffectiveness of traditional education in these situations, it is essential to teach road awareness through simulations and games to young traffic users. Children can be immersed in the virtual environment and feel the sense of presence in the real environment. Children can experience many risky scenarios in the virtual world and learn how to avoid them. This paper introduces a serious game to educate children about traffic laws. Seven crucial traffic regulations for children when crossing the street are presented by two scenarios. The learning process is carried out during playing in three stages: pretest, test, and posttest. Children receive feedback regarding the suitability of their traffic behavior using the coins they acquire during the game. Both of the game's scenarios were found to be effective at helping players understand traffic laws, according to the pretest and posttest results. © 2023 IEEE.

关键词： Roads and streets

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：