Table 3 Comparison of different infrared small target detection methods on the MDvsFA-cGAN dataset.
From: Deep asymmetric extraction and aggregation for infrared small target detection
| Â | Method | \(IoU(\times 10^{-2})\) | \(P_{d}(\times 10^{-2})\) | \(F_{a}(\times 10^{-5})\) | Time (s) | FLOPs (G) | Params (M) |
|---|---|---|---|---|---|---|---|
Model-Driven | Top-Hat10 | 4.59 | 61.87 | 151.09 | 0.0288 | – | – |
Max-Median11 | 3.06 | 54.67 | 222.76 | 0.0176 | – | – | |
WSLCM29 | 12.73 | 92.81 | 259.35 | 14.1103 | – | – | |
TLCM30 | 7.61 | 77.69 | 357.62 | 2.9738 | – | – | |
IPI15 | 17.04 | 76.97 | 3.06 | 0.5691 | – | – | |
NRAM31 | 10.01 | 54.67 | 2.56 | 4.2193 | – | – | |
RIPT16 | 13.26 | 92.08 | 138.24 | 1.0926 | – | – | |
PSTNN32 | 16.64 | 69.06 | 3.52 | 0.7275 | – | – | |
MSLSTIPT33 | 5.12 | 48.92 | 2.21 | 0.083 | – | – | |
Data-Driven | U-Net20 | 45.78 | 85.08 | 15.04 | 0.069 | 0.31 | 1.6 |
ACM5 | 46.65 | 84.92 | 9.03 | 0.078 | 0.30 | 1.6 | |
ALC12 | 46.35 | 85.16 | 17.32 | 0.072 | 0.41 | 1.44 | |
DNANet22 | 43.96 | 78.41 | 10.43 | 0.064 | 10.91 | 18.7 | |
ISTDU-Net13 | 41.96 | 71.94 | 15.25 | 0.072 | 6.08 | 11.3 | |
DAEA(S=3) | 47.13 | 85.61 | 23.59 | 0.064 | 1.96 | 1.2 | |
DAEA(S=4) | 47.28 | 87.05 | 20.41 | 0.066 | 2.43 | 1.6 | |
DAEA(S=5) | 48.37 | 86.33 | 10.67 | 0.072 | 2.90 | 2.0 | |
DAEA(S=6) | 48.79 | 84.89 | 7.34 | 0.076 | 3.35 | 2.4 |
