In the learning process of the metric-based meta-learning. there are some problems. such as the lack of prior knowledge acquired due to the distribution of scarce data. the interference of weakly related or unrelated features extracted from a single-view sample. https://poolsproductscanadas.shop/product-category/hayward-h135-parts/