Classification of COVID-CT Images Utilizing Four Types of Deep Convolutional Neural Networks
Апстракт
In this paper a method is presented for the
classification of COVID-CT (CT_COVID, CT_NonCOVID)
image data set. Four different types of deep convolutional
neural networks are proposed, two with the architecture
resembling the VGGNet, one resembling the LeNet-5 and
one using transfer learning. In addition, neural networks
utilized the following techniques: decay, dropout and batch
normalization. Since we needed to combat a significantly
small dataset, we used data augmentation in order to
transform and expand our dataset. Moreover,
juxtapositions were made when observing the results given
by these four neural networks, as well as the affect made by
two different optimizers. The training of the neural
networks was done using small batches with a binary cross
entropy loss function, in order to achieve an up to scratch
classification accuracy.
Кључне речи:
deep learning / convolutional neural networks / image classification / data augmentation / batch normalization / COVID-CT dataset / dropout / transfer learningИзвор:
THE FIFTH INTERNATIONAL CONFERENCE “MECHANICAL ENGINEERING IN XXI CENTURY” - MASING 2020 PROCEEDINGS -UNIVERSITY OF NIŠ FACULTY OF MECHANICAL ENGINEERING IN NIŠ, December 9-10., 2020, MECHATRONICS AND CONTROL pp. 201-206Издавач:
- Faculty of Mechanical Engineering in Niš
- Prof. Dr Nenad T. Pavlović, Dean
Финансирање / пројекти:
- MISSION4.0 - Deep Machine Learning and Swarm Intelligence-Based Optimization Algorithms for Control and Scheduling of Cyber-Physical Systems in Industry 4.0 (RS-ScienceFundRS-AI-6523109)
- Министарство науке, технолошког развоја и иновација Републике Србије, институционално финансирање - 200105 (Универзитет у Београду, Машински факултет) (RS-MESTD-inst-2020-200105)
- Развој методологија за повећање радне способности, поузданости и енергетске ефикасности машинских система у енергетици (RS-MESTD-Technological Development (TD or TR)-35029)
Колекције
Институција/група
Mašinski fakultetTY - CONF AU - Laban, Lara AU - Vesović, Mitra PY - 2020 UR - https://machinery.mas.bg.ac.rs/handle/123456789/4508 AB - In this paper a method is presented for the classification of COVID-CT (CT_COVID, CT_NonCOVID) image data set. Four different types of deep convolutional neural networks are proposed, two with the architecture resembling the VGGNet, one resembling the LeNet-5 and one using transfer learning. In addition, neural networks utilized the following techniques: decay, dropout and batch normalization. Since we needed to combat a significantly small dataset, we used data augmentation in order to transform and expand our dataset. Moreover, juxtapositions were made when observing the results given by these four neural networks, as well as the affect made by two different optimizers. The training of the neural networks was done using small batches with a binary cross entropy loss function, in order to achieve an up to scratch classification accuracy. PB - Faculty of Mechanical Engineering in Niš PB - Prof. Dr Nenad T. Pavlović, Dean C3 - THE FIFTH INTERNATIONAL CONFERENCE “MECHANICAL ENGINEERING IN XXI CENTURY” - MASING 2020 PROCEEDINGS -UNIVERSITY OF NIŠ FACULTY OF MECHANICAL ENGINEERING IN NIŠ, December 9-10. T1 - Classification of COVID-CT Images Utilizing Four Types of Deep Convolutional Neural Networks EP - 206 SP - MECHATRONICS AND CONTROL pp. 201 UR - https://hdl.handle.net/21.15107/rcub_machinery_4508 ER -
@conference{ author = "Laban, Lara and Vesović, Mitra", year = "2020", abstract = "In this paper a method is presented for the classification of COVID-CT (CT_COVID, CT_NonCOVID) image data set. Four different types of deep convolutional neural networks are proposed, two with the architecture resembling the VGGNet, one resembling the LeNet-5 and one using transfer learning. In addition, neural networks utilized the following techniques: decay, dropout and batch normalization. Since we needed to combat a significantly small dataset, we used data augmentation in order to transform and expand our dataset. Moreover, juxtapositions were made when observing the results given by these four neural networks, as well as the affect made by two different optimizers. The training of the neural networks was done using small batches with a binary cross entropy loss function, in order to achieve an up to scratch classification accuracy.", publisher = "Faculty of Mechanical Engineering in Niš, Prof. Dr Nenad T. Pavlović, Dean", journal = "THE FIFTH INTERNATIONAL CONFERENCE “MECHANICAL ENGINEERING IN XXI CENTURY” - MASING 2020 PROCEEDINGS -UNIVERSITY OF NIŠ FACULTY OF MECHANICAL ENGINEERING IN NIŠ, December 9-10.", title = "Classification of COVID-CT Images Utilizing Four Types of Deep Convolutional Neural Networks", pages = "206-MECHATRONICS AND CONTROL pp. 201", url = "https://hdl.handle.net/21.15107/rcub_machinery_4508" }
Laban, L.,& Vesović, M.. (2020). Classification of COVID-CT Images Utilizing Four Types of Deep Convolutional Neural Networks. in THE FIFTH INTERNATIONAL CONFERENCE “MECHANICAL ENGINEERING IN XXI CENTURY” - MASING 2020 PROCEEDINGS -UNIVERSITY OF NIŠ FACULTY OF MECHANICAL ENGINEERING IN NIŠ, December 9-10. Faculty of Mechanical Engineering in Niš., MECHATRONICS AND CONTROL pp. 201-206. https://hdl.handle.net/21.15107/rcub_machinery_4508
Laban L, Vesović M. Classification of COVID-CT Images Utilizing Four Types of Deep Convolutional Neural Networks. in THE FIFTH INTERNATIONAL CONFERENCE “MECHANICAL ENGINEERING IN XXI CENTURY” - MASING 2020 PROCEEDINGS -UNIVERSITY OF NIŠ FACULTY OF MECHANICAL ENGINEERING IN NIŠ, December 9-10.. 2020;:MECHATRONICS AND CONTROL pp. 201-206. https://hdl.handle.net/21.15107/rcub_machinery_4508 .
Laban, Lara, Vesović, Mitra, "Classification of COVID-CT Images Utilizing Four Types of Deep Convolutional Neural Networks" in THE FIFTH INTERNATIONAL CONFERENCE “MECHANICAL ENGINEERING IN XXI CENTURY” - MASING 2020 PROCEEDINGS -UNIVERSITY OF NIŠ FACULTY OF MECHANICAL ENGINEERING IN NIŠ, December 9-10. (2020):MECHATRONICS AND CONTROL pp. 201-206, https://hdl.handle.net/21.15107/rcub_machinery_4508 .