Method and device for explainable few-shot image classification

摘要

The present invention discloses a method and apparatus for classifying a fractional shot image that can be described. According to the present invention, as an explanatory small-shot image classification apparatus, a latent space generator that generates a latent variable for the real image by inputting a real image as an input and an image generator that generates a virtual artificial image by inputting a random latent variable as an input an adversarial generative neural network including a generator and a discriminator that calculates and outputs a probability that the artificial image belongs to the real image; and classifying whether or not different first and second images have the same object using the latent space generator on which learning has been completed, and synthesizing latent variables for each of the first and second images, There is provided an explanatory fractional-shot image classification apparatus including a fractional-shot learning unit that generates an image for qualitative evaluation.