Complexity and accuracy analysis of common artificial neural networks on pedestrian detection
With the development of computer version, deep learning and artificial neural networks approaches like SPP-net, Faster-RCNN and YOLO are proposed. This paper compares them in terms of efficiency and effectiveness. By analyzing the network architecture, SPP-net is more complex than Faster-RCNN and YO...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2018-01-01
|
| Series: | MATEC Web of Conferences |
| Online Access: | https://doi.org/10.1051/matecconf/201823201003 |
| Summary: | With the development of computer version, deep learning and artificial neural networks approaches like SPP-net, Faster-RCNN and YOLO are proposed. This paper compares them in terms of efficiency and effectiveness. By analyzing the network architecture, SPP-net is more complex than Faster-RCNN and YOLO. By analyzing the experiments, SPP-net and Faster-RCNN are more accurate than YOLO in static detection while opposite in real-time system. Therefore, in real-time pedestrian detection situation, YOLO can perform better. In static pedestrian detection situation, Faster-RCNN or SPP-net can perform better. |
|---|---|
| ISSN: | 2261-236X |