One-Shot Vision, AI for Effortless Identification of Individuals with and Without Glasses

Facial recognition is a challenging task, especially in the wild where there exist a substantial amount of variation in lighting, pose, and expression. One-shot knowledge acquisition is a form of machine learning that allows a model to learn to recognize a new class Upon observing just a single instance of that class. the proposed work of a one-shot learning strategy for facial recognition system that is specifically designed to detect persons with or without spectacles. The proposed work approach has the capability to achieve high accuracy on a limited dataset of labeled images, and possesses the capacity to adjust to unfamiliar images that were not encountered in prior training stages. In this work, One-shot learning aims to recognize individuals based on a limited number of examples with and without specs, making it suitable for scenarios where acquiring a large dataset is challenging.