Sample mapping-based few-sample underwater acoustic target identification method

• Main Authors: XUE YANGTAO, LI XUELONG, CUI XIAODONG, WANG ZHEN, ZHU PEICAN
• Format: Patent
• Language: Chinese; English
• Published: 27.06.2023
• Subjects: ACOUSTICS; CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; MUSICAL INSTRUMENTS; PHYSICS; SPEECH ANALYSIS OR SYNTHESIS; SPEECH OR AUDIO CODING OR DECODING; SPEECH OR VOICE PROCESSING; SPEECH RECOGNITION

The invention provides a few-sample underwater acoustic target recognition method based on sample mapping. Comprising the following steps: step 1, performing a classification task on a base data set to train a feature extraction module for extracting features of a new class; 2, performing feature extraction on the samples of the new classes by using a trained feature extraction module, calculating the average feature of each class, mapping the samples on the base data set to a type space formed by the new classes according to the similarity of the samples to the center feature of each class, and marking pseudo labels on the samples; and step 3, performing fine tuning on the model by using the base data set sample with the pseudo tag and the support set sample to fit feature recognition of a new class. Compared with other models, the method provided by the invention has higher accuracy when the underwater target is detected under the condition of few samples, and meanwhile, a better effect can be obtained unde