A Derivation of vicinal kernel functions) is derived as Z Q(z

Each class contains 600 images of size 84 84. These classes are split into 64, 16, and 20 classes for meta-training, meta-validation, and meta-testing respectively [28]. CUB contains 200 classes with a total of 11,788 images of size 84 84. Following previous works [5], the base, validation, and novel split are 100, 50, and 50 classes respectively. CIFAR-FS is a variant of the CIFAR-100 dataset used for few-shot classification. It contains 100 classes, each with 600 images of 32 32 pixels. The classes are randomly split into 64, 16, and 20 for meta-training, meta-validation, and meta-testing respectively.