A Review of Underwater Target Recognition Based on Deep Learning Models

Underwater detection technology is widely used in ocean exploration nowadays, but it has some problems such as noisy detection signals, difficult feature extraction and complicated calculation. With the excellent performance of artificial intelligence in the field of signal processing, deep learning technology has been paid more and more attention by researchers. This paper summarizes the application and development of deep learning model in underwater target detection. Including Long Short-Term Memory, Transformer model, Convolutional Neural Networks, Residual network, etc. The shortcomings of existing deep learning models in underwater target recognition are analyzed, and improved methods such as small sample detection, multi-modal fusion algorithm and edge computing are proposed. Finally, it summarizes and predicts the trend of technological development in this field.