To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
Ideep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Innovation and investment strategies to intensify the potential modernization and to increase the competitiveness of microeconomic systems |
|
Author |
Svitlana Tulchynska, Olha Vovk, Olha Popelo, Stanislav Saloid, Olena Kostiunik |
| Citation |
Vol. 21 No. 7 pp. 161-168 |
|
Abstract |
Within the article, strategic guidelines for the modernization of microeconomic systems are identified. Modernization levels of the potential implementation are formalized for enterprises: contractile, extensive technical, technological, progressive, adaptive, steady, intensive, creative, absolute and leader modernization. This allowed to specify the directions and tasks of the enterprise modernization at different management levels. Accordingly, the conditions and criteria for selecting resource tools are set. It is proved that the strategies of the potential modernization of enterprises must be carried out at four main management levels: first, at the enterprise level; secondly, for a particular type of product / service; third, by functional directions of modernization of separate spheres of the enterprise activity or responsibility, fourth, at the level of structural units of the enterprise. It is substantiated that in the processes due to the activation of the potential modernization, the resources are transformed into the results of the innovation implementation and the investment strategies modernization. A system of tasks for the corporate strategies implementation in order to modernize microeconomic systems has been formed. Key vectors of the activation determine the nature and properties of investment resources and necessary innovations to enhance the modernization potential. Therefore, the system of innovation and investment strategies’ modernization, based on the vector and resource provision of the modernization process, is specified: |
|
Keywords |
Investment and Innovation Strategies, Microeconomic System, Modernization, Investment Activity, Innovative Development. |
|
URL |
|
|
Title |
Innovation and investment strategies to intensify the potential modernization and to increase the competitiveness of microeconomic systems |
|
Author |
Svitlana Tulchynska, Olha Vovk, Olha Popelo, Stanislav Saloid, Olena Kostiunik |
| Citation |
Vol. 21 No. 7 pp. 161-168 |
|
Abstract |
Within the article, strategic guidelines for the modernization of microeconomic systems are identified. Modernization levels of the potential implementation are formalized for enterprises: contractile, extensive technical, technological, progressive, adaptive, steady, intensive, creative, absolute and leader modernization. This allowed to specify the directions and tasks of the enterprise modernization at different management levels. Accordingly, the conditions and criteria for selecting resource tools are set. It is proved that the strategies of the potential modernization of enterprises must be carried out at four main management levels: first, at the enterprise level; secondly, for a particular type of product / service; third, by functional directions of modernization of separate spheres of the enterprise activity or responsibility, fourth, at the level of structural units of the enterprise. It is substantiated that in the processes due to the activation of the potential modernization, the resources are transformed into the results of the innovation implementation and the investment strategies modernization. A system of tasks for the corporate strategies implementation in order to modernize microeconomic systems has been formed. Key vectors of the activation determine the nature and properties of investment resources and necessary innovations to enhance the modernization potential. Therefore, the system of innovation and investment strategies’ modernization, based on the vector and resource provision of the modernization process, is specified: |
|
Keywords |
Investment and Innovation Strategies, Microeconomic System, Modernization, Investment Activity, Innovative Development. |
|
URL |
|
|
Title |
Innovation and investment strategies to intensify the potential modernization and to increase the competitiveness of microeconomic systems |
|
Author |
Svitlana Tulchynska, Olha Vovk, Olha Popelo, Stanislav Saloid, Olena Kostiunik |
| Citation |
Vol. 21 No. 7 pp. 161-168 |
|
Abstract |
Within the article, strategic guidelines for the modernization of microeconomic systems are identified. Modernization levels of the potential implementation are formalized for enterprises: contractile, extensive technical, technological, progressive, adaptive, steady, intensive, creative, absolute and leader modernization. This allowed to specify the directions and tasks of the enterprise modernization at different management levels. Accordingly, the conditions and criteria for selecting resource tools are set. It is proved that the strategies of the potential modernization of enterprises must be carried out at four main management levels: first, at the enterprise level; secondly, for a particular type of product / service; third, by functional directions of modernization of separate spheres of the enterprise activity or responsibility, fourth, at the level of structural units of the enterprise. It is substantiated that in the processes due to the activation of the potential modernization, the resources are transformed into the results of the innovation implementation and the investment strategies modernization. A system of tasks for the corporate strategies implementation in order to modernize microeconomic systems has been formed. Key vectors of the activation determine the nature and properties of investment resources and necessary innovations to enhance the modernization potential. Therefore, the system of innovation and investment strategies’ modernization, based on the vector and resource provision of the modernization process, is specified: |
|
Keywords |
Investment and Innovation Strategies, Microeconomic System, Modernization, Investment Activity, Innovative Development. |
|
URL |
|
|
Title |
Innovation and investment strategies to intensify the potential modernization and to increase the competitiveness of microeconomic systems |
|
Author |
Svitlana Tulchynska, Olha Vovk, Olha Popelo, Stanislav Saloid, Olena Kostiunik |
| Citation |
Vol. 21 No. 7 pp. 161-168 |
|
Abstract |
Within the article, strategic guidelines for the modernization of microeconomic systems are identified. Modernization levels of the potential implementation are formalized for enterprises: contractile, extensive technical, technological, progressive, adaptive, steady, intensive, creative, absolute and leader modernization. This allowed to specify the directions and tasks of the enterprise modernization at different management levels. Accordingly, the conditions and criteria for selecting resource tools are set. It is proved that the strategies of the potential modernization of enterprises must be carried out at four main management levels: first, at the enterprise level; secondly, for a particular type of product / service; third, by functional directions of modernization of separate spheres of the enterprise activity or responsibility, fourth, at the level of structural units of the enterprise. It is substantiated that in the processes due to the activation of the potential modernization, the resources are transformed into the results of the innovation implementation and the investment strategies modernization. A system of tasks for the corporate strategies implementation in order to modernize microeconomic systems has been formed. Key vectors of the activation determine the nature and properties of investment resources and necessary innovations to enhance the modernization potential. Therefore, the system of innovation and investment strategies’ modernization, based on the vector and resource provision of the modernization process, is specified: |
|
Keywords |
Investment and Innovation Strategies, Microeconomic System, Modernization, Investment Activity, Innovative Development. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Innovation and investment strategies to intensify the potential modernization and to increase the competitiveness of microeconomic systems |
|
Author |
Svitlana Tulchynska, Olha Vovk, Olha Popelo, Stanislav Saloid, Olena Kostiunik |
| Citation |
Vol. 21 No. 7 pp. 161-168 |
|
Abstract |
Within the article, strategic guidelines for the modernization of microeconomic systems are identified. Modernization levels of the potential implementation are formalized for enterprises: contractile, extensive technical, technological, progressive, adaptive, steady, intensive, creative, absolute and leader modernization. This allowed to specify the directions and tasks of the enterprise modernization at different management levels. Accordingly, the conditions and criteria for selecting resource tools are set. It is proved that the strategies of the potential modernization of enterprises must be carried out at four main management levels: first, at the enterprise level; secondly, for a particular type of product / service; third, by functional directions of modernization of separate spheres of the enterprise activity or responsibility, fourth, at the level of structural units of the enterprise. It is substantiated that in the processes due to the activation of the potential modernization, the resources are transformed into the results of the innovation implementation and the investment strategies modernization. A system of tasks for the corporate strategies implementation in order to modernize microeconomic systems has been formed. Key vectors of the activation determine the nature and properties of investment resources and necessary innovations to enhance the modernization potential. Therefore, the system of innovation and investment strategies’ modernization, based on the vector and resource provision of the modernization process, is specified: |
|
Keywords |
Investment and Innovation Strategies, Microeconomic System, Modernization, Investment Activity, Innovative Development. |
|
URL |
|
|
Title |
Innovation and investment strategies to intensify the potential modernization and to increase the competitiveness of microeconomic systems |
|
Author |
Svitlana Tulchynska, Olha Vovk, Olha Popelo, Stanislav Saloid, Olena Kostiunik |
| Citation |
Vol. 21 No. 7 pp. 161-168 |
|
Abstract |
Within the article, strategic guidelines for the modernization of microeconomic systems are identified. Modernization levels of the potential implementation are formalized for enterprises: contractile, extensive technical, technological, progressive, adaptive, steady, intensive, creative, absolute and leader modernization. This allowed to specify the directions and tasks of the enterprise modernization at different management levels. Accordingly, the conditions and criteria for selecting resource tools are set. It is proved that the strategies of the potential modernization of enterprises must be carried out at four main management levels: first, at the enterprise level; secondly, for a particular type of product / service; third, by functional directions of modernization of separate spheres of the enterprise activity or responsibility, fourth, at the level of structural units of the enterprise. It is substantiated that in the processes due to the activation of the potential modernization, the resources are transformed into the results of the innovation implementation and the investment strategies modernization. A system of tasks for the corporate strategies implementation in order to modernize microeconomic systems has been formed. Key vectors of the activation determine the nature and properties of investment resources and necessary innovations to enhance the modernization potential. Therefore, the system of innovation and investment strategies’ modernization, based on the vector and resource provision of the modernization process, is specified: |
|
Keywords |
Investment and Innovation Strategies, Microeconomic System, Modernization, Investment Activity, Innovative Development. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
|
|
Title |
Food Detection by Fine-Tuning Pre-trained Convolutional Neural Network Using Noisy Labels |
|
Author |
Shroog Alshomrani, Lina Aljoudi, Banan Aljabri, Sarah Al-Shareef |
| Citation |
Vol. 21 No. 7 pp. 182-190 |
|
Abstract |
Deep learning is an advanced technology for large-scale data analysis, with numerous promising cases like image processing, object detection and significantly more. It becomes customarily to use transfer learning and fine-tune a pre-trained CNN model for most image recognition tasks. Having people taking photos and tag themselves provides a valuable resource of in-data. However, these tags and labels might be noisy as people who annotate these images might not be experts. This paper aims to explore the impact of noisy labels on fine-tuning pre-trained CNN models. Such effect is measured on a food recognition task using Food101 as a benchmark. Four pre-trained CNN models are included in this study: InceptionV3, VGG19, MobileNetV2 and DenseNet121. Symmetric label noise will be added with different ratios. In all cases, models based on DenseNet121 outperformed the other models. When noisy labels were introduced to the data, the performance of all models degraded almost linearly with the amount of added noise. |
|
Keywords |
deep learning, food image detection, symmetric label noise, convolutional neural networks, transfer learning. |
|
URL |
