Pub Date : 2022-09-01DOI: 10.1109/AI4I54798.2022.00007
Alberto Cruz, Stephanie Magana, David Greco, L. Bellis, A. Luvisi
Almond Leaf Scorch Disease (ALSD) poses a signfficant threat to almond production worldwide. Deep learning algorithms have potential to enable growers of all scales to identify infected trees using photos captured with a smart camera. This approach diagnoses faster than humans without requiring expert knowledge, does not require third-party laboratory testing (PCR), and has higher accuracy than multispectral satillite imaging. Data was collected for this work by long-term observation of Prunus dulcis L. for symptoms and validated ALSD with PCR testing. 515 images were collected. We experimented with five pre-trained convolutional neural networks: DenseNet 201, Inception V3, ResNet 101 V2, VGG 19, and Xception. DenseNet201 demonstrates that it is possible to detect ALSD versus healthy control, Red Leaf Blotch, and various other diseases with an 88.72% accuracy. These results show promise for cheap and fast detection of the disease. Future work will focus on imaging and detection of the disease in vivo.
{"title":"Detection of Almond Leaf Scorch with Artificial Intelligence for the Agriculture Industry","authors":"Alberto Cruz, Stephanie Magana, David Greco, L. Bellis, A. Luvisi","doi":"10.1109/AI4I54798.2022.00007","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00007","url":null,"abstract":"Almond Leaf Scorch Disease (ALSD) poses a signfficant threat to almond production worldwide. Deep learning algorithms have potential to enable growers of all scales to identify infected trees using photos captured with a smart camera. This approach diagnoses faster than humans without requiring expert knowledge, does not require third-party laboratory testing (PCR), and has higher accuracy than multispectral satillite imaging. Data was collected for this work by long-term observation of Prunus dulcis L. for symptoms and validated ALSD with PCR testing. 515 images were collected. We experimented with five pre-trained convolutional neural networks: DenseNet 201, Inception V3, ResNet 101 V2, VGG 19, and Xception. DenseNet201 demonstrates that it is possible to detect ALSD versus healthy control, Red Leaf Blotch, and various other diseases with an 88.72% accuracy. These results show promise for cheap and fast detection of the disease. Future work will focus on imaging and detection of the disease in vivo.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124823762","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00020
Young-Seob Jeong, Eunjin Kim, JunHa Hwang, M. E. Mswahili, Youngjin Kim
Since transformer has appeared, there were many studies that proposed variants of some representative language models (e.g., Bidirectional Encoder Representations from Transformers (BERT) [1] and Generative Pre-Training (GPT) series [2]). Huge language models are appearing recently (e.g., Chinchilla [3], Megatron LM), whereas there are studies of domain-specific (or language-specific) language models. For example, BioBERT for bio-informatics [4], SwahBERT for Swahili language [5], and FinBERT for financial domain [6]. Without doubt, statistics must be one of the domains with many collected data (e.g., reports of statistics). Pre-trained language model for the statistic domain will probably deliver much performance improvement in down-stream tasks such as industry code classification and job code classification, and more accurate system for the code classification tasks will contribute to better national statistics and taxation. Indeed, many countries are trying to develop such system, and this paper summarizes some relevant findings and provides suggestions to develop language models for statistics domain.
{"title":"Language Model for Statistics Domain","authors":"Young-Seob Jeong, Eunjin Kim, JunHa Hwang, M. E. Mswahili, Youngjin Kim","doi":"10.1109/AI4I54798.2022.00020","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00020","url":null,"abstract":"Since transformer has appeared, there were many studies that proposed variants of some representative language models (e.g., Bidirectional Encoder Representations from Transformers (BERT) [1] and Generative Pre-Training (GPT) series [2]). Huge language models are appearing recently (e.g., Chinchilla [3], Megatron LM), whereas there are studies of domain-specific (or language-specific) language models. For example, BioBERT for bio-informatics [4], SwahBERT for Swahili language [5], and FinBERT for financial domain [6]. Without doubt, statistics must be one of the domains with many collected data (e.g., reports of statistics). Pre-trained language model for the statistic domain will probably deliver much performance improvement in down-stream tasks such as industry code classification and job code classification, and more accurate system for the code classification tasks will contribute to better national statistics and taxation. Indeed, many countries are trying to develop such system, and this paper summarizes some relevant findings and provides suggestions to develop language models for statistics domain.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116157419","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00009
Mouzhi Ge, G. Pilato, Fabio Persia, D. D’Auria
Nowadays, recommender systems are increasingly being exploited in many industrial applications, including virtual museums and movie streaming platforms. In the last few years, some new perspectives provided by research paradigms such as deep learning or quantum computing, have arisen. As a result, this paper identifies four new perspectives on recommender systems: e-health, tourism, deep-learning-based, and recommender systems exploiting quantum computing. After discussing them, the paper provides the current state of the art and highlights the possible future directions for industries.
{"title":"New Perspectives on Recommender Systems for Industries","authors":"Mouzhi Ge, G. Pilato, Fabio Persia, D. D’Auria","doi":"10.1109/AI4I54798.2022.00009","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00009","url":null,"abstract":"Nowadays, recommender systems are increasingly being exploited in many industrial applications, including virtual museums and movie streaming platforms. In the last few years, some new perspectives provided by research paradigms such as deep learning or quantum computing, have arisen. As a result, this paper identifies four new perspectives on recommender systems: e-health, tourism, deep-learning-based, and recommender systems exploiting quantum computing. After discussing them, the paper provides the current state of the art and highlights the possible future directions for industries.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126394776","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00011
M. E. Mswahili, Junha Hwang, Young-Seob Jeong, Youngjin Kim
In this study, we implemented graph neural network (GNN) methods to forecast in vitro inhibitory bioactivity or pharmacological concentration of chemical compounds against severe acute respiratory syndrome (SARS) coronaviruses from the graph representation amongst the compounds (i.e., nodes) and their respective features(i.e., node features) obtained by RDKit tool from their respectively SMILES (Simplified MolecularInput Line-Entry System), and we compared GNN models by experiments with our graph data of 375 nodes with 44,475 edges or links. This was done in response to the severe and significant consequences of the ongoing Coronavirus disease 2019 (COVID-19) disease. As a result, we discovered that implemented models, simple graph convolution (SGC), and graph convolution network (GCN) performed significantly well with comparable performance.
{"title":"Graph Neural Network Models for Chemical Compound Activeness Prediction For COVID-19 Drugs Discovery using Lipinski’s Descriptors","authors":"M. E. Mswahili, Junha Hwang, Young-Seob Jeong, Youngjin Kim","doi":"10.1109/AI4I54798.2022.00011","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00011","url":null,"abstract":"In this study, we implemented graph neural network (GNN) methods to forecast in vitro inhibitory bioactivity or pharmacological concentration of chemical compounds against severe acute respiratory syndrome (SARS) coronaviruses from the graph representation amongst the compounds (i.e., nodes) and their respective features(i.e., node features) obtained by RDKit tool from their respectively SMILES (Simplified MolecularInput Line-Entry System), and we compared GNN models by experiments with our graph data of 375 nodes with 44,475 edges or links. This was done in response to the severe and significant consequences of the ongoing Coronavirus disease 2019 (COVID-19) disease. As a result, we discovered that implemented models, simple graph convolution (SGC), and graph convolution network (GCN) performed significantly well with comparable performance.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127055790","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00008
Kévin Cédric Guyard, Michel Deriaz
Recommendation systems are becoming more and more present in our daily lives, whether it is for suggesting items to buy, movies to watch or music to listen. They can be used in a large number of contexts. In this paper, we propose the use of a recommendation system in the context of a recruitment platform. The use of the recommendation system allows to obtain precise profile recommendations based on the competences of a candidate to meet the stated requirements and to avoid companies to have to perform a very time-consuming manual sorting of the candidates. Thus, this paper presents the context in which we propose this recommendation system, the data preprocessing, the general approach based on a hybrid Content Based Filtering (CBF) and Similarity Index (SI) system, as well as the means implemented to reduce the computational cost of such a system with the increasing evolution of the platform.
{"title":"A scalable recommendation system approach for a companies - seniors matching","authors":"Kévin Cédric Guyard, Michel Deriaz","doi":"10.1109/AI4I54798.2022.00008","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00008","url":null,"abstract":"Recommendation systems are becoming more and more present in our daily lives, whether it is for suggesting items to buy, movies to watch or music to listen. They can be used in a large number of contexts. In this paper, we propose the use of a recommendation system in the context of a recruitment platform. The use of the recommendation system allows to obtain precise profile recommendations based on the competences of a candidate to meet the stated requirements and to avoid companies to have to perform a very time-consuming manual sorting of the candidates. Thus, this paper presents the context in which we propose this recommendation system, the data preprocessing, the general approach based on a hybrid Content Based Filtering (CBF) and Similarity Index (SI) system, as well as the means implemented to reduce the computational cost of such a system with the increasing evolution of the platform.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"14 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114031185","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00018
Bastian Scharnagl, Christian Groth
EEG classification is a promising approach to facilitate the life of handicapped people and to generate future human-computer-interfaces. In this paper we want to compare the effectiveness of current state of the art deep learning techniques for EEG classification. Therefore, we applied different approaches on various datasets and did a crosscomparison of the results in order to get more knowledge on the generalization capabilities. Additionally, we created a new EEG dataset and made it available for further research.
{"title":"Evaluation of different deep learning approaches for EEG classification","authors":"Bastian Scharnagl, Christian Groth","doi":"10.1109/AI4I54798.2022.00018","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00018","url":null,"abstract":"EEG classification is a promising approach to facilitate the life of handicapped people and to generate future human-computer-interfaces. In this paper we want to compare the effectiveness of current state of the art deep learning techniques for EEG classification. Therefore, we applied different approaches on various datasets and did a crosscomparison of the results in order to get more knowledge on the generalization capabilities. Additionally, we created a new EEG dataset and made it available for further research.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"-1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115010367","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 : 2022-09-01DOI: 10.1109/ai4i54798.2022.00025
Richard Yarnell, Daniel Brignac, Yanjie Fu, R. Demara
Neural network-based object detection has many important applications but requires a vast amount of training data. In applications where training data may be scarce, data augmentation techniques can be used to expand the training set. This paper explores the performance of such techniques on You Only Look Once Version 5 (YOLOv5).
基于神经网络的目标检测有许多重要的应用，但需要大量的训练数据。在训练数据稀缺的应用中，可以使用数据增强技术来扩展训练集。本文探讨了这些技术在You Only Look Once Version 5 (YOLOv5)上的性能。
{"title":"Utilization of Data Augmentation Techniques to Enhance Learning with Sparse Datasets","authors":"Richard Yarnell, Daniel Brignac, Yanjie Fu, R. Demara","doi":"10.1109/ai4i54798.2022.00025","DOIUrl":"https://doi.org/10.1109/ai4i54798.2022.00025","url":null,"abstract":"Neural network-based object detection has many important applications but requires a vast amount of training data. In applications where training data may be scarce, data augmentation techniques can be used to expand the training set. This paper explores the performance of such techniques on You Only Look Once Version 5 (YOLOv5).","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129176044","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00024
Raphael Wallsberger, Tim Dieter Eberhardt, Paul-Albert Bartlau, Maurice Lucas Dörnte, Tim Lukas Schröter, S. Matzka
In this paper, a new visualization-based method for understanding industrial machine learning models trained on high dimensional data is proposed. For a view that includes all dimensions, a dimensionality reduction towards a 2D projection, the UMAP method, is used. With the TreeSHAP algorithm the most important features for each machine condition are identified, visualized and evaluated. A closer look at different data points of the most important features provides more information about the behavior of the model. In addition, this knowledge is used to derive a class-optimized 2D visualization to increase trustworthiness of individual classification results for domain experts.
{"title":"Explainable Artificial Intelligence for a high dimensional condition monitoring application using the SHAP Method","authors":"Raphael Wallsberger, Tim Dieter Eberhardt, Paul-Albert Bartlau, Maurice Lucas Dörnte, Tim Lukas Schröter, S. Matzka","doi":"10.1109/AI4I54798.2022.00024","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00024","url":null,"abstract":"In this paper, a new visualization-based method for understanding industrial machine learning models trained on high dimensional data is proposed. For a view that includes all dimensions, a dimensionality reduction towards a 2D projection, the UMAP method, is used. With the TreeSHAP algorithm the most important features for each machine condition are identified, visualized and evaluated. A closer look at different data points of the most important features provides more information about the behavior of the model. In addition, this knowledge is used to derive a class-optimized 2D visualization to increase trustworthiness of individual classification results for domain experts.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"444-445 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114732026","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 : 2022-09-01DOI: 10.1109/ai4i54798.2022.00006
{"title":"Message from the AI4I 2022 Program Co-Chairs","authors":"","doi":"10.1109/ai4i54798.2022.00006","DOIUrl":"https://doi.org/10.1109/ai4i54798.2022.00006","url":null,"abstract":"","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133523880","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 : 2022-09-01DOI: 10.1109/AI4I54798.2022.00013
A. Samad, Towneda Akhter Prema
Automatic Real-Time Vehicle License Plate recognition systemsface a wide variety of challenges to accurately reading the number sequences present in license plates when deployed in the real world. The uncertainty of the real-world license plate recognition system is addressed in this research and proposes a novel method of filtering out inaccurate results based on recurrence-based sorting of the best readable character sequences. The problems of unconstrained real-world real-time license plate recognition have been discussed thoroughly throughout the paper and redefined the problem statement towards attaining the best readability instead of best visibility. An end-to-end cascade neural network architecture has been developed to capture license plate readings in real time. Then the novel recurrence-based iteration method was introduced to sort the Top 1 of the best reading of a unique license plate tracked across the frame where character sequence reading accuracy has been improved by up to 54% when combined with the introduced garbage factor filtering method. The experimental evidence indicated that the traditional confidence-based sorting is prone to failure due to unconstrained real-world uncertainties and accuracy is thoroughly compared with our novel method to illustrate novelty. The system has been deployed in the real world on an embedded system with a substantial amount of vehicle traffic for testing and validation.
{"title":"Recurrence sorting method for improved accuracy of unconstrained fast-moving vehicle license plate recognition system","authors":"A. Samad, Towneda Akhter Prema","doi":"10.1109/AI4I54798.2022.00013","DOIUrl":"https://doi.org/10.1109/AI4I54798.2022.00013","url":null,"abstract":"Automatic Real-Time Vehicle License Plate recognition systemsface a wide variety of challenges to accurately reading the number sequences present in license plates when deployed in the real world. The uncertainty of the real-world license plate recognition system is addressed in this research and proposes a novel method of filtering out inaccurate results based on recurrence-based sorting of the best readable character sequences. The problems of unconstrained real-world real-time license plate recognition have been discussed thoroughly throughout the paper and redefined the problem statement towards attaining the best readability instead of best visibility. An end-to-end cascade neural network architecture has been developed to capture license plate readings in real time. Then the novel recurrence-based iteration method was introduced to sort the Top 1 of the best reading of a unique license plate tracked across the frame where character sequence reading accuracy has been improved by up to 54% when combined with the introduced garbage factor filtering method. The experimental evidence indicated that the traditional confidence-based sorting is prone to failure due to unconstrained real-world uncertainties and accuracy is thoroughly compared with our novel method to illustrate novelty. The system has been deployed in the real world on an embedded system with a substantial amount of vehicle traffic for testing and validation.","PeriodicalId":345427,"journal":{"name":"2022 5th International Conference on Artificial Intelligence for Industries (AI4I)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129597282","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：