Figure 8
From: Few-shot learning for joint model in underwater acoustic target recognition
Comparison of the joint model performance under five cases on three extreme few-shot datasets. (a) Results on 1 shot dataset; (b) results on 5 shot dataset; (c) results on 20-shot dataset. The X-axis coordinate represents the number of stages utilized in training process, and the Y-axis represents the accuracy of the joint model. Each shape point represents a stage combination. The number near the shape points represents the accuracy of the joint model.
