MambaDSF: Multi-Scale SSM with Dilated Feature Fusion for Sonar Small Target Detection
MambaDSF is a hybrid framework for detecting small targets in sonar imaging, overcoming challenges like low contrast and scale ambiguity. It combines a MambaEFP backbone with a Dilate Fusion Mamba encoder and specialized losses to achieve state-of-the-art performance (91.5% mAP50). This method effectively captures both local echoes and global context for superior small target detection.