Pub Date : 2024-06-10DOI: 10.15625/1813-9663/18058
Luong Tran Thi
Block ciphers in general, Substitution-Permutation Network (SPN) block ciphers in particular are cryptographic fields widely applied today. AES is an SPN block cipher used in many security applications. However, there are many strong attacks on block ciphers as linear attacks, differential attacks, and algebraic attacks which are challenging for cryptographers. Therefore, the research to improve the security of block ciphers in general and AES, in particular, is a topic of great interest today. Along with security, the issue of the execution cost of block ciphers is also crucial in practice. In this paper, we clarify the role of the MDS matrix in increasing the branch number of the diffusion layer of the block ciphers, thereby improving the security of the block ciphers. We propose a method improving the security of the AES block cipher by changing the Mixcolumn transformation of AES using execution-efficient MDS matrices of size 4, 8, or 16. We present a method to find a new diffusion matrix of modified AES block ciphers from which to evaluate the number of fixed points and coefficient of fixed points of the modified AES diffusion layers. In addition, we prove the branch number of the modified AES diffusion layers with MDS matrices of sizes 8, and 16. Then we also analyze the security, statistical standards and execution speed of modified AES block ciphers generated from those MDS matrices. The results show that our proposed method can significantly improve the security of the AES block cipher.
{"title":"PROVING THE SECURITY OF AES BLOCK CIPHER BASED ON MODIFIED MIXCOLUMN","authors":"Luong Tran Thi","doi":"10.15625/1813-9663/18058","DOIUrl":"https://doi.org/10.15625/1813-9663/18058","url":null,"abstract":"Block ciphers in general, Substitution-Permutation Network (SPN) block ciphers in particular are cryptographic fields widely applied today. AES is an SPN block cipher used in many security applications. However, there are many strong attacks on block ciphers as linear attacks, differential attacks, and algebraic attacks which are challenging for cryptographers. Therefore, the research to improve the security of block ciphers in general and AES, in particular, is a topic of great interest today. Along with security, the issue of the execution cost of block ciphers is also crucial in practice. In this paper, we clarify the role of the MDS matrix in increasing the branch number of the diffusion layer of the block ciphers, thereby improving the security of the block ciphers. We propose a method improving the security of the AES block cipher by changing the Mixcolumn transformation of AES using execution-efficient MDS matrices of size 4, 8, or 16. We present a method to find a new diffusion matrix of modified AES block ciphers from which to evaluate the number of fixed points and coefficient of fixed points of the modified AES diffusion layers. In addition, we prove the branch number of the modified AES diffusion layers with MDS matrices of sizes 8, and 16. Then we also analyze the security, statistical standards and execution speed of modified AES block ciphers generated from those MDS matrices. The results show that our proposed method can significantly improve the security of the AES block cipher.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":" 19","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141365079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-25DOI: 10.15625/1813-9663/19018
Dinh Duc Luong, Vuong Quang Phuong, Hoang Do Thanh Tung
The exponential growth of bioinformatics in the healthcare domain has revolutionized our understanding of DNA, proteins, and other biomolecular entities. This remarkable progress has generated an overwhelming volume of data, necessitating big data technologies for efficient storage and indexing. While big data technologies like Hadoop offer substantial support for big XML file storage, the challenges of indexing data sizes and XPath query performance persist. To enhance the efficiency of XPath queries and address the data size problem, a novel approach that is derived from the spatial indexing method of the R-tre family. The proposed method is to modify the structure of leaf nodes in the indexing tree to preserve XML-sibling connections. Then, new algorithms for constructing the new tree structure and processing sibling queries better are introduced. Experimental results demonstrate the superior efficiency of sibling XPath queries with reduced data sizes for indexing, while other XPath queries exhibit notable performance improvements. This research contributes to the development of more effective indexing methods for managing and querying large XML datasets in bioinformatics applications, ultimately advancing biomedical research and healthcare initiatives.
生物信息学在医疗保健领域的飞速发展彻底改变了我们对 DNA、蛋白质和其他生物分子实体的认识。这一令人瞩目的进步产生了大量数据，需要大数据技术来进行高效存储和索引。虽然 Hadoop 等大数据技术为大型 XML 文件存储提供了大量支持，但索引数据规模和 XPath 查询性能方面的挑战依然存在。为了提高 XPath 查询的效率并解决数据大小问题，一种源自 R-tre 系列空间索引方法的新方法应运而生。所提出的方法是修改索引树中叶节点的结构，以保留 XML 同胞连接。然后，介绍了构建新的树结构和更好地处理同胞查询的新算法。实验结果表明，在索引数据量减少的情况下，同胞 XPath 查询的效率更高，而其他 XPath 查询的性能也有显著提高。这项研究有助于开发更有效的索引方法，用于管理和查询生物信息学应用中的大型 XML 数据集，最终推动生物医学研究和医疗保健计划。
{"title":"AN IMPROVED INDEXING METHOD FOR QUERYING BIG XML FILES","authors":"Dinh Duc Luong, Vuong Quang Phuong, Hoang Do Thanh Tung","doi":"10.15625/1813-9663/19018","DOIUrl":"https://doi.org/10.15625/1813-9663/19018","url":null,"abstract":"The exponential growth of bioinformatics in the healthcare domain has revolutionized our understanding of DNA, proteins, and other biomolecular entities. This remarkable progress has generated an overwhelming volume of data, necessitating big data technologies for efficient storage and indexing. While big data technologies like Hadoop offer substantial support for big XML file storage, the challenges of indexing data sizes and XPath query performance persist. To enhance the efficiency of XPath queries and address the data size problem, a novel approach that is derived from the spatial indexing method of the R-tre family. The proposed method is to modify the structure of leaf nodes in the indexing tree to preserve XML-sibling connections. Then, new algorithms for constructing the new tree structure and processing sibling queries better are introduced. Experimental results demonstrate the superior efficiency of sibling XPath queries with reduced data sizes for indexing, while other XPath queries exhibit notable performance improvements. This research contributes to the development of more effective indexing methods for managing and querying large XML datasets in bioinformatics applications, ultimately advancing biomedical research and healthcare initiatives.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"1 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139157596","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-25DOI: 10.15625/1813-9663/18122
Luan Ngo Dinh, Hiếu Lê Ngọc, Long Quoc Phan
Multilingual Visual Question Answering (mVQA) is an extremely challenging task which needs to answer a question given in different languages and take the context in an image. This problem can only be addressed by the combination of Natural Language Processing and Computer Vision. In this paper, we propose applying a jointly developed model to the task of multilingual visual question answering. Specifically, we conduct experiments on a multimodal sequence-to-sequence transformer model derived from the T5 encoder-decoder architecture. Text tokens and Vision Transformer (ViT) dense image embeddings are inputs to an encoder then we used a decoder to automatically anticipate discrete text tokens. We achieved the F1-score of 0.4349 on the private test set and ranked 2nd in the EVJVQA task at the VLSP shared task 2022. For reproducing the result, the code can be found at https://github.com/DinhLuan14/VLSP2022-VQA-OhYeah
{"title":"OHYEAH AT VLSP2022-EVJVQA CHALLENGE: A JOINTLY LANGUAGE-IMAGE MODEL FOR MULTILINGUAL VISUAL QUESTION ANSWERING","authors":"Luan Ngo Dinh, Hiếu Lê Ngọc, Long Quoc Phan","doi":"10.15625/1813-9663/18122","DOIUrl":"https://doi.org/10.15625/1813-9663/18122","url":null,"abstract":"Multilingual Visual Question Answering (mVQA) is an extremely challenging task which needs to answer a question given in different languages and take the context in an image. This problem can only be addressed by the combination of Natural Language Processing and Computer Vision. In this paper, we propose applying a jointly developed model to the task of multilingual visual question answering. Specifically, we conduct experiments on a multimodal sequence-to-sequence transformer model derived from the T5 encoder-decoder architecture. Text tokens and Vision Transformer (ViT) dense image embeddings are inputs to an encoder then we used a decoder to automatically anticipate discrete text tokens. We achieved the F1-score of 0.4349 on the private test set and ranked 2nd in the EVJVQA task at the VLSP shared task 2022. For reproducing the result, the code can be found at https://github.com/DinhLuan14/VLSP2022-VQA-OhYeah","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"21 7","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139157603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-21DOI: 10.15625/1813-9663/18680
Pham Viet Anh, Vu Duc Thi, Nguyen Ngoc Cuong
Attribute reduction, or attribute selection in the decision table, is a fundamental problem of rough set theory. Currently, many scientists are interested in and developing these issues. Unfortunately, most studies focus mainly on the complete decision table. On incomplete decision tables, researchers have proposed tolerance relations and designed attribute reduction algorithms based on different measures. However, these algorithms only return a reduct and do not preserve information in the decision tables. This paper will propose an efficient method to determine entire reducts of incomplete decision tables according to the relational database approach. In the complex case, this algorithm has exponential computational complexity. However, this algorithm has polynomial computational complexity in the different cases of databases.
{"title":"A NOVEL ALGORITHM FOR FINDING ALL REDUCTS IN THE INCOMPLETE DECISION TABLE","authors":"Pham Viet Anh, Vu Duc Thi, Nguyen Ngoc Cuong","doi":"10.15625/1813-9663/18680","DOIUrl":"https://doi.org/10.15625/1813-9663/18680","url":null,"abstract":"Attribute reduction, or attribute selection in the decision table, is a fundamental problem of rough set theory. Currently, many scientists are interested in and developing these issues. Unfortunately, most studies focus mainly on the complete decision table. On incomplete decision tables, researchers have proposed tolerance relations and designed attribute reduction algorithms based on different measures. However, these algorithms only return a reduct and do not preserve information in the decision tables. This paper will propose an efficient method to determine entire reducts of incomplete decision tables according to the relational database approach. In the complex case, this algorithm has exponential computational complexity. However, this algorithm has polynomial computational complexity in the different cases of databases.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"18 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139253590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-21DOI: 10.15625/1813-9663/18236
Van Thang Nguyen, Thanh Long Luong, Huan Vu
Emotional speech synthesis is a challenging task in speech processing. To build an emotional Text-to-speech (TTS) system, one would need to have a quality emotional dataset of the target speaker. However, collecting such data is difficult, sometimes even impossible. This paper presents our approach that addresses the problem of transplanting a source speaker's emotional expression to a target speaker, one of the Vietnamese Language and Speech Processsing (VLSP) 2022 TTS tasks. Our approach includes a complete data pre-processing pipeline and two training algorithms. We first train a source speaker's expressive TTS model, then adapt the voice characteristics for the target speaker. Empirical results have shown the efficacy of our method in generating the expressive speech of a speaker under a limited training data regime.
{"title":"THE VNPT-IT EMOTION TRANSPLANTATION APPROACH FOR VLSP 2022","authors":"Van Thang Nguyen, Thanh Long Luong, Huan Vu","doi":"10.15625/1813-9663/18236","DOIUrl":"https://doi.org/10.15625/1813-9663/18236","url":null,"abstract":"Emotional speech synthesis is a challenging task in speech processing. To build an emotional Text-to-speech (TTS) system, one would need to have a quality emotional dataset of the target speaker. However, collecting such data is difficult, sometimes even impossible. This paper presents our approach that addresses the problem of transplanting a source speaker's emotional expression to a target speaker, one of the Vietnamese Language and Speech Processsing (VLSP) 2022 TTS tasks. Our approach includes a complete data pre-processing pipeline and two training algorithms. We first train a source speaker's expressive TTS model, then adapt the voice characteristics for the target speaker. Empirical results have shown the efficacy of our method in generating the expressive speech of a speaker under a limited training data regime.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"83 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139251953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-21DOI: 10.15625/1813-9663/18043
Dat Tien Nguyen, Chau Ngoc Ha, Ha Thanh Thi Hoang, Truong Nhat Nguyen, Tuyet Ngoc Huynh, Hai Thanh Nguyen
Practicing sports is an activity that helps people maintain and improve their health, enhance memory and concentration, reduce anxiety and stress, and train teamwork and leadership ability. With the development of science and technology, artificial intelligence in sports has become increasingly popular with the public and brings many benefits. In particular, many applications help people track and evaluate athletes' achievements in competitions. This study extracts images from Taekwondo videos and generates skeleton data from frames using the Fast Forward Moving Picture Experts Group (FFMPEG) technique using MoveNet. After that, we use deep learning architectures such as Long Short-Term Memory Networks, Convolutional Long Short-Term Memory, and Long-term Recurrent Convolutional Networks to perform the poses classification tasks in Taegeuk in Jang lessons. This work presents two approaches. The first approach uses a sequence skeleton extracted from the image by Movenet. Second, we use sequence images to train using video classification architecture. Finally, we recognize poses in sports lessons using skeleton data to remove noise in the image, such as background and extraneous objects behind the exerciser. As a result, our proposed method has achieved promising performance in pose classification tasks in an introductory Taekwondo lesson.
{"title":"TAEKWONDO POSE ESTIMATION WITH DEEP LEARNING ARCHITECTURES ON ONE-DIMENSIONAL AND TWO-DIMENSIONAL DATA","authors":"Dat Tien Nguyen, Chau Ngoc Ha, Ha Thanh Thi Hoang, Truong Nhat Nguyen, Tuyet Ngoc Huynh, Hai Thanh Nguyen","doi":"10.15625/1813-9663/18043","DOIUrl":"https://doi.org/10.15625/1813-9663/18043","url":null,"abstract":"Practicing sports is an activity that helps people maintain and improve their health, enhance memory and concentration, reduce anxiety and stress, and train teamwork and leadership ability. With the development of science and technology, artificial intelligence in sports has become increasingly popular with the public and brings many benefits. In particular, many applications help people track and evaluate athletes' achievements in competitions. This study extracts images from Taekwondo videos and generates skeleton data from frames using the Fast Forward Moving Picture Experts Group (FFMPEG) technique using MoveNet. After that, we use deep learning architectures such as Long Short-Term Memory Networks, Convolutional Long Short-Term Memory, and Long-term Recurrent Convolutional Networks to perform the poses classification tasks in Taegeuk in Jang lessons. This work presents two approaches. The first approach uses a sequence skeleton extracted from the image by Movenet. Second, we use sequence images to train using video classification architecture. Finally, we recognize poses in sports lessons using skeleton data to remove noise in the image, such as background and extraneous objects behind the exerciser. As a result, our proposed method has achieved promising performance in pose classification tasks in an introductory Taekwondo lesson.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"59 18","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139253167","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-10-06DOI: 10.15625/1813-9663/18144
Dang Ngan Ha, Hieu Trung Huynh
Plant species recognition plays an important role in agriculture, the pharmaceutical industry, and conservation. The traditional approaches may take days and have difficulties for non-experts. Several computer vision-based models have been proposed, which can partially assist and speed up the plant recognition process. Thanks to the development of data collection and computational systems, the models based on machine learning have considerably improved their performance in the last decades. In this paper, we present a model for plant recognition in Southeast Asia based on the high-resolution network. The evaluation is carried out on a public dataset consisting of 26 different species in Southeast Asia. It shows high accuracy in recognition.
{"title":"A PLANT RECOGNITION APPROACH USING HIGH RESOLUTION NETWORK","authors":"Dang Ngan Ha, Hieu Trung Huynh","doi":"10.15625/1813-9663/18144","DOIUrl":"https://doi.org/10.15625/1813-9663/18144","url":null,"abstract":"Plant species recognition plays an important role in agriculture, the pharmaceutical industry, and conservation. The traditional approaches may take days and have difficulties for non-experts. Several computer vision-based models have been proposed, which can partially assist and speed up the plant recognition process. Thanks to the development of data collection and computational systems, the models based on machine learning have considerably improved their performance in the last decades. In this paper, we present a model for plant recognition in Southeast Asia based on the high-resolution network. The evaluation is carried out on a public dataset consisting of 26 different species in Southeast Asia. It shows high accuracy in recognition.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134946454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-26DOI: 10.15625/1813-9663/18157
Nguyen, Ngan Luu-Thuy, Nguyen, Nghia Hieu, Vo, Duong T. D, Tran, Khanh Quoc, Van Nguyen, Kiet
Visual Question Answering (VQA) is a challenging task of natural language processing (NLP) and computer vision (CV), attracting significant attention from researchers. English is a resource-rich language that has witnessed various developments in datasets and models for visual question answering. Visual question answering in other languages also would be developed for resources and models. In addition, there is no multilingual dataset targeting the visual content of a particular country with its own objects and cultural characteristics. To address the weakness, we provide the research community with a benchmark dataset named EVJVQA, including 33,000+ pairs of question-answer over three languages: Vietnamese, English, and Japanese, on approximately 5,000 images taken from Vietnam for evaluating multilingual VQA systems or models. EVJVQA is used as a benchmark dataset for the challenge of multilingual visual question answering at the 9th Workshop on Vietnamese Language and Speech Processing (VLSP 2022). This task attracted 62 participant teams from various universities and organizations. In this article, we present details of the organization of the challenge, an overview of the methods employed by shared-task participants, and the results. The highest performances are 0.4392 in F1-score and 0.4009 in BLUE on the private test set. The multilingual QA systems proposed by the top 2 teams use ViT for the pre-trained vision model and mT5 for the pre-trained language model, a powerful pre-trained language model based on the transformer architecture. EVJVQA is a challenging dataset that motivates NLP and CV researchers to further explore the multilingual models or systems for visual question answering systems.
{"title":"EVJVQA CHALLENGE: MULTILINGUAL VISUAL QUESTION ANSWERING","authors":"Nguyen, Ngan Luu-Thuy, Nguyen, Nghia Hieu, Vo, Duong T. D, Tran, Khanh Quoc, Van Nguyen, Kiet","doi":"10.15625/1813-9663/18157","DOIUrl":"https://doi.org/10.15625/1813-9663/18157","url":null,"abstract":"Visual Question Answering (VQA) is a challenging task of natural language processing (NLP) and computer vision (CV), attracting significant attention from researchers. English is a resource-rich language that has witnessed various developments in datasets and models for visual question answering. Visual question answering in other languages also would be developed for resources and models. In addition, there is no multilingual dataset targeting the visual content of a particular country with its own objects and cultural characteristics. To address the weakness, we provide the research community with a benchmark dataset named EVJVQA, including 33,000+ pairs of question-answer over three languages: Vietnamese, English, and Japanese, on approximately 5,000 images taken from Vietnam for evaluating multilingual VQA systems or models. EVJVQA is used as a benchmark dataset for the challenge of multilingual visual question answering at the 9th Workshop on Vietnamese Language and Speech Processing (VLSP 2022). This task attracted 62 participant teams from various universities and organizations. In this article, we present details of the organization of the challenge, an overview of the methods employed by shared-task participants, and the results. The highest performances are 0.4392 in F1-score and 0.4009 in BLUE on the private test set. The multilingual QA systems proposed by the top 2 teams use ViT for the pre-trained vision model and mT5 for the pre-trained language model, a powerful pre-trained language model based on the transformer architecture. EVJVQA is a challenging dataset that motivates NLP and CV researchers to further explore the multilingual models or systems for visual question answering systems.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134903242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-22DOI: 10.15625/1813-9663/18259
Khang Nguyen
Drones are increasingly used in various application domains including surveillance, agriculture, delivery, search and rescue missions. Object detection in aerial images (captured by drones) gradually gains more interest in computer vision community. However, research activities are still very few in this area due to numerous challenges such as top-view angle, small-scale object, diverse directions, and data imbalance. In this paper, we investigate different data augmentation techniques. Furthermore, we propose combining data augmentation methods to further enhance the performance of the state-of-the-art object detection methods. Extensive experiments on two datasets, namely, AERIAU, and XDUAV, demonstrate that the combination of random cropped and vertical flipped data boosts the performance of object detectors on aerial images.
{"title":"DATA AUGMENTATION ANALYSIS OF VEHICLE DETECTION IN AERIAL IMAGES","authors":"Khang Nguyen","doi":"10.15625/1813-9663/18259","DOIUrl":"https://doi.org/10.15625/1813-9663/18259","url":null,"abstract":"Drones are increasingly used in various application domains including surveillance, agriculture, delivery, search and rescue missions. Object detection in aerial images (captured by drones) gradually gains more interest in computer vision community. However, research activities are still very few in this area due to numerous challenges such as top-view angle, small-scale object, diverse directions, and data imbalance. In this paper, we investigate different data augmentation techniques. Furthermore, we propose combining data augmentation methods to further enhance the performance of the state-of-the-art object detection methods. Extensive experiments on two datasets, namely, AERIAU, and XDUAV, demonstrate that the combination of random cropped and vertical flipped data boosts the performance of object detectors on aerial images.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136100800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-22DOI: 10.15625/1813-9663/18015
Truong Thi Hong Thuy, Nguyen Hoang Phuong
Traditional Vietnamese Medicine (TVM) is based on the experiences of thousands of years of Vietnamese people in the struggle against diseases; therefore, TVM is very important in the medical system of Vietnam. In this paper, we propose a novel model of an expert system for diagnosing disease syndromes and treating traditional Vietnamese medicine. In this model, the knowledge base consists of IF-THEN rules, in which the antecedent of a rule is an elementary conjunction of propositions and negated propositions. The inference mechanism for the diagnosis of disease syndromes and treatment of traditional Vietnamese medicine applies Abelian group operations. A comparison of the inference of our model with the fuzzy max-min inferences shows that our model can have very similar rules whose contributions sum up to high weight. On the other hand, in our model, a rule with a negative weight may diminish an effect of a rule with a good weight. This feature is absent in the systems with fuzzy max-min inferences. We have built rule patterns for the diagnosis of about 50 disease syndromes and their treatment by Herbs and Acupuncture with the cooperation of practitioners of Oriental Traditional Medicine in Vietnam. Some examples of databases and the rules for disease syndrome differentiation and treatment by herbal medicine and Acupuncture are shown. Finally, some conclusions and future works are given.
{"title":"A NOVEL APPROACH TO MODELLING A DIAGNOSIS AND TREATMENT OF TRADITIONAL VIETNAMESE MEDICINE","authors":"Truong Thi Hong Thuy, Nguyen Hoang Phuong","doi":"10.15625/1813-9663/18015","DOIUrl":"https://doi.org/10.15625/1813-9663/18015","url":null,"abstract":"Traditional Vietnamese Medicine (TVM) is based on the experiences of thousands of years of Vietnamese people in the struggle against diseases; therefore, TVM is very important in the medical system of Vietnam. In this paper, we propose a novel model of an expert system for diagnosing disease syndromes and treating traditional Vietnamese medicine. In this model, the knowledge base consists of IF-THEN rules, in which the antecedent of a rule is an elementary conjunction of propositions and negated propositions. The inference mechanism for the diagnosis of disease syndromes and treatment of traditional Vietnamese medicine applies Abelian group operations. A comparison of the inference of our model with the fuzzy max-min inferences shows that our model can have very similar rules whose contributions sum up to high weight. On the other hand, in our model, a rule with a negative weight may diminish an effect of a rule with a good weight. This feature is absent in the systems with fuzzy max-min inferences. We have built rule patterns for the diagnosis of about 50 disease syndromes and their treatment by Herbs and Acupuncture with the cooperation of practitioners of Oriental Traditional Medicine in Vietnam. Some examples of databases and the rules for disease syndrome differentiation and treatment by herbal medicine and Acupuncture are shown. Finally, some conclusions and future works are given.","PeriodicalId":15444,"journal":{"name":"Journal of Computer Science and Cybernetics","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136100798","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：