Table 7.
Comparisons of the number of parameters, Floating-point Operations(FLOPs), inference time and Frame Per Second(FPS).
| NETWORKS | Params (M) |
FLOPs (G) |
Inference time (msec) |
FPS |
|---|---|---|---|---|
| DeepLab V3+ | 40.37 | 11.71 | 46.52 | 21.50 |
| SEGNet | 29.44 | 40.20 | 152.23 | 6.57 |
| ENet | 0.35 | 0.55 | 69.23 | 14.44 |
| UNet | 13.40 | 31.13 | 123.69 | 8.08 |
| UNeXt | 1.47 | 0.58 | 21.39 | 46.74 |
| MFRWF-CWF | 3.14 | 24.00 | 74.97 | 13.33 |
| Ours | 0.11 | 0.24 | 18.14 | 55.10 |
The best results for each metric are highlighted in bold.