Analysis of the Application Efficiency of TensorFlow and PyTorch in Convolutional Neural Network
Ovidiu-Constantin Novac,
Mihai Cristian Chirodea,
Cornelia Mihaela Novac,
Nicu Bizon,
Mihai Oproescu,
Ovidiu Petru Stan,
Cornelia Emilia Gordan
Affiliations
Ovidiu-Constantin Novac
Department of Computers and Information Technology, Electrical Engineering and Information Technology Faculty, University of Oradea, 410087 Oradea, Romania
Mihai Cristian Chirodea
Department of Computers and Information Technology, Electrical Engineering and Information Technology Faculty, University of Oradea, 410087 Oradea, Romania
Cornelia Mihaela Novac
Department of Electrical Engineering, Electrical Engineering and Information Technology Faculty, University of Oradea, 410087 Oradea, Romania
Nicu Bizon
Department of Electronics, Computers and Electrical Engineering, Faculty of Electronics, Telecommunication, and Computer Science, University of Pitesti, 110040 Pitesti, Romania
Mihai Oproescu
Department of Electronics, Computers and Electrical Engineering, Faculty of Electronics, Telecommunication, and Computer Science, University of Pitesti, 110040 Pitesti, Romania
Ovidiu Petru Stan
Department of Automation, Faculty of Automation and Computer Science, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania
Cornelia Emilia Gordan
Department of Electronics and Telecommunications, Electrical Engineering and Information Technology Faculty, University of Oradea, 410087 Oradea, Romania
In this paper, we present an analysis of important aspects that arise during the development of neural network applications. Our aim is to determine if the choice of library can impact the system’s overall performance, either during training or design, and to extract a set of criteria that could be used to highlight the advantages and disadvantages of each library under consideration. To do so, we first extracted the previously mentioned aspects by comparing two of the most popular neural network libraries—PyTorch and TensorFlow—and then we performed an analysis on the obtained results, with the intent of determining if our initial hypothesis was correct. In the end, the results of the analysis are gathered, and an overall picture of what tasks are better suited for what library is presented.
