Search

Underwater object detection is a prerequisite for underwater robots to achieve autonomous operation and ocean exploration. However, poor imaging quality, harsh underwater environments and concealed underwater targets greatly aggravate the difficulty of underwater object detection. In order to reduce underwater background interference and improve underwater object perception, we propose a multi-dimensional, multi-functional and multi-level attention module (mDFLAM). The multi-dimensional strategy first enhances the robustness of attention application by collecting valuable information in different target dimensions. The multi-functional strategy further improves the flexibility of attention calibration by capturing the importance of channel semantic information and the dependence of spatial location information. The multi-level strategy finally enriches the diversity of attention perception by extracting the intrinsic information under different receptive fields. In pre-processing and post-processing stages, cross-splitting and cross-linking stimulate the synergistic calibration advantage of multi-dimensional and multi-functional attention by redistributing channel dimensions and restoring feature states. In the attention calibration stage, adaptive fusion stimulates the synergistic calibration advantage of multi-level attention by assigning learnable parameters. In order to meet the high-precision and real-time requirements for underwater object detection, we integrate the plug-and-play mDFLAM into YOLO detectors. The full-port embedding further strengthens the semantic information expression by improving the feature fusion quality between scales. In underwater detection tasks, ablation and comparison experiments demonstrate the rationality and effectiveness of our attention design. In other detection tasks, our work shows good robustness and generalization. [ABSTRACT FROM AUTHOR]

Marine object detection has become increasingly important in intelligent underwater robot. Because of color cast and blur in underwater images, features directly extracted from backbone networks usually lack interesting and discriminative characters, that affects performance on marine object detection. To this end, this paper proposes a novel refined marine object detector with attention-based spatial pyramid pooling networks and bidirectional feature fusion strategy to relieve the weakening of features and address marine object detection issues. Firstly, an attention-based spatial pyramid pooling network named as SA-SPPN is proposed to enrich interesting information and extend receptive field on original features extracted from backbone network. Based on enhanced multiple level features, the bidirectional feature fusion strategy is designed to fuse different level features and generate robust feature maps for detection. Specifically, the top-down connection could transfer semantic information from high-level features to enhance low-level features. The bottom-up pathway could extend resolution of high-level features. Furthermore, the cross-layer connections are integrated into both top-down passway and bottom-up passway to carry out multiple branch fusion. On bounding boxes regression phase, the distance-IoU loss is adopted to improve regression speed and accuracy. Finally, this paper conducts series experiments on underwater image datasets and URPC datasets to detect marine objects. The experimental results reveal that our approach could achieve impressive performance and reach 79.64% mAP on underwater image datasets, 79.31% mAP on URPC2019 datasets and 79.93% mAP on URPC2020 datasets, respectively. For standard object detection, the proposed algorithm also could realize notable performance and get 81.9% mAP on PASCAL VOC datasets. [ABSTRACT FROM AUTHOR]