SW_YoloX_N: YoloX-Nano with sliding Windows
Video not available.
Rendering of new sequences is currently deactivated due to heavy load.
Rendering of new sequences is currently deactivated due to heavy load.
Rendering of new sequences is currently deactivated due to heavy load.
Rendering of new sequences is currently deactivated due to heavy load.
Rendering of new sequences is currently deactivated due to heavy load.
Rendering of new sequences is currently deactivated due to heavy load.
Rendering of new sequences is currently deactivated due to heavy load.
Short name:
SW_YoloX_N
Description:
A Lightweight version of Shift Window-YOLOX (SW-YOLOX)
Reference:
C.-Y. Tsai, R.-Y. Wang and Y.-C. Chiu, “SW-YOLOX: A Real-Time Pedestrian Detection System Based on Sliding Window-Mixed Attention Mechanism in YOLOX Lightweight Network,” Neurocomputing, under review.
Last submitted:
July 04, 2023 (9 months ago)
Published:
July 04, 2023 at 16:20:28 CET
Submissions:
1
Project page / code:
n/a
Open source:
No
Hardware:
2.4 GHZ，24 Core
Runtime:
46.4 Hz
Benchmark performance:
| Sequence | AP | MODA | MODP | FAF | TP | FP | FN | Rcll | Prcn | F1 |
| MOT17Det | 0.82 | 82.5 | 82.2 | 1.3 | 102,022 | 7,502 | 12,542 | 89.1 | 93.2 | 91.1 |
Detailed performance:
| Sequence | AP | MODA | MODP | FAF | TP | FP | FN | Rcll | Prcn | F1 |
| MOT17-01 | 0.63 | 63.9 | 82.3 | 0.3 | 2,783 | 119 | 1,383 | 66.8 | 95.9 | 78.7 |
| MOT17-03 | 0.91 | 91.5 | 82.7 | 3.2 | 67,170 | 4,840 | 952 | 98.6 | 93.3 | 95.9 |
| MOT17-06 | 0.81 | 73.9 | 79.7 | 0.5 | 7,005 | 541 | 1,744 | 80.1 | 92.8 | 86.0 |
| MOT17-07 | 0.81 | 82.0 | 81.2 | 1.0 | 8,102 | 513 | 1,158 | 87.5 | 94.0 | 90.7 |
| MOT17-08 | 0.91 | 79.5 | 85.0 | 1.5 | 5,520 | 951 | 225 | 96.1 | 85.3 | 90.4 |
| MOT17-12 | 0.73 | 70.9 | 83.8 | 0.2 | 3,974 | 151 | 1,420 | 73.7 | 96.3 | 83.5 |
| MOT17-14 | 0.54 | 53.9 | 77.4 | 0.5 | 7,468 | 387 | 5,660 | 56.9 | 95.1 | 71.2 |
Raw data:
