Published 2020 | Version v1
Journal article

Comparative Performance Analysis of Neural Network Real-Time Object Detections in Different Implementations

  • 1. OOO «Videointellect», Skolkovo Innovation Centre, 42 Bolshoy boulevard, 143026 Moscow (Russian Federation)
  • 2. Institute of Experimental Physics, Slovak Academy of Sciences Watsonova 47, 04001 Košice (Slovakia)
  • 3. The Laboratory of Information Technologies, JINR, 6 Joliot-Curie, 141980 Dubna, Moscow Region (Russian Federation)

Description

The performance of neural networks is one of the most important topics in the field of computer vision. In this work, we analyze the speed of object detection using the well-known YOLOv3 neural network architecture in different frameworks under different hardware requirements. We obtain results, which allow us to formulate preliminary qualitative conclusions about the feasibility of various hardware scenarios to solve tasks in real-time environments.

Availability note (English)

Available from https://www.epj-conferences.org/articles/epjconf/pdf/2020/02/epjconf_mmcp2019_02020.pdf; https://doaj.org/article/2661962922de47598c12dc31d48a9e6f

Additional details

Publishing Information

Journal Title
EPJ. Web of Conferences
Journal Volume
226
Journal Page Range
vp.
ISSN
2100-014X

Conference

Title
International Conference on Mathematical Modeling and Computational Physics
Acronym
MMCP 2019
Dates
1-5 Jul 2019
Place
Stara Lesna (Slovakia)

INIS

Country of Publication
France
Country of Input or Organization
France
INIS RN
53115745
Subject category
S97: MATHEMATICAL METHODS AND COMPUTING;
Resource subtype / Literary indicator
Conference
Descriptors DEI
COMPUTER ARCHITECTURE; COMPUTERS; DETECTION; IMPLEMENTATION; NEURAL NETWORKS; PERFORMANCE; VISION