Table 1 The ablation experiment results for object detection performance on the VEDAI dataset.
From: Super Mamba feature enhancement framework for small object detection
RFAConv | SAM | SE | FEM | P | R | mAP@0.5 | PSNR | mAP@0.95 | Pare(M) |
|---|---|---|---|---|---|---|---|---|---|
× | × | × | × | 0.714 | 0.720 | 0.700 | 42.76 | 0.416 | 8.26 |
√ | × | × | × | 0.740 | 0.673 | 0.722 | 43.69 | 0.423 | 9.52 |
× | √ | × | × | 0.781 | 0.679 | 0.747 | 44.66 | 0.454 | 9.13 |
× | × | √ | × | 0.7145 | 0.7202 | 0.6645 | 41.31 | 0.407 | 7.52 |
× | × | × | √ | 0.721 | 0.745 | 0.769 | 44.65 | 0.473 | 18.2 |
√ | √ | √ | √ | 0.914 | 0.919 | 0.923 | 45.70 | 0.778 | 17.6 |