Perception enhanced hybrid network for underwater object detection

Table 1. Quantitative comparison of different algorithms on the TrashCan dataset
View table
View in Article
Table 1. Quantitative comparison of different algorithms on the TrashCan dataset
方法 $m A P_{0.5 ∶ 0.95} / %$ $A P_{50} / %$
Faster R-CNN 31.2 55.3
AutoAssign 30.7 51.1
Deforable DETR 24.1 43.4
PVT 31.4 52.7
ERL-Net 37.0 58.9
Boosting R-CNN 36.8 57.6
RoIAttn 32.6 57.2
本文 37.8 61.8

Table 2. Quantitative comparison of different algorithms on the WPBB dataset
View table
View in Article
Table 2. Quantitative comparison of different algorithms on the WPBB dataset
方法 $m A P_{0.5 : 0.95} / %$ $A P_{50} / %$
Faster R-CNN 75.7 98.7
AutoAssign 75.9 98.8
Deforable DETR 73.2 98.3
PVT 79.5 98.5
ERL-Net 78.5 97.1
Boosting R-CNN 70.2 88.1
RoIAttn 79.7 98.5
本文 82.0 98.9

Table 3. Ablation experiments on the TrashCan dataset
View table
View in Article
Table 3. Ablation experiments on the TrashCan dataset
卷积注意力增强模块目标感知增强双阶段检测头 $m A P_{0.5 ∶ 0.95} / %$ $A P_{50} / %$
× × 31.4 52.7
√ × 33.7 56.6
× √ 37.4 60.2
√ √ 37.8 61.8

Tools

Get Citation

Copy Citation Text

Tingting YAO, Ning LI, Yu ZHANG. Perception enhanced hybrid network for underwater object detection[J]. Optics and Precision Engineering, 2025, 33(8): 1303

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites