Fig 1.
Structure of the ReDet.
Fig 2.
The structure of the M-ReDet network.
Fig 3.
The structure of the SOPM module.
Table 1.
Experimental results of the M-Bottleneck configuration.
Fig 4.
Structure of the SFRM module.
Fig 5.
The number of each type ship in the FAIR1M(ship) dataset.
Table 2.
Experimental results for different loss functions.
Table 3.
Experimental details table.
Fig 6.
Loss function curves for the training.
Fig 7.
Fig 7 is attributed to the DOTA open-source database and is available from the DOTA database (URL (s): https://captain-whu.github.io/DOTA/dataset.html).
Fig 8.
The number of various types in the DOTA.
Fig 9.
Examples of the FAIR1M(ship) dataset.
Fig 9 is attributed to the FAIR1M open-source database and is available from the FAIR1M database (URL (s): https://gaofen-challenge.com/benchmark).
Fig 10.
Detection results of the M-ReDet on the DOTA dataset.
Fig 10 is attributed to the DOTA open-source database and is available from the DOTA database (URL (s): https://captain-whu.github.io/DOTA/dataset.html).
Table 4.
Results of the DOTA dataset comparison experiment.
Table 5.
Detection results of various types of remote sensing objects.
Fig 11.
Training results of each algorithm on the DOTA dataset.
Fig 12.
Ship object detection and fine-grained recognition results.
Fig 12 is attributed to the FAIR1M open-source database and is available from the FAIR1M database (URL (s): https://gaofen-challenge.com/benchmark).
Table 6.
Comparison experiment results on the FAIR1M(ship) dataset.
Table 7.
Training results for various types of ships.
Table 8.
Detection and fine-grained recognition results for 9 types of ships.
Fig 13.
Comparison of training results on the FAIR1M(ship) dataset.
Table 9.
Ablation experiments of the M-ReDet.
Fig 14.
Ship detection and fine-grained results.
(a)Detection and fine-grained results of the ReDet; (b)Detection and fine-grained results of the M-ReDet. Fig 14 is attributed to the FAIR1M open-source database and is available from the FAIR1M database (URL (s): https://gaofen-challenge.com/benchmark).