The Large-Scale Korean Influencer Dataset (KoIn) presents over 100,000 images collected from Instagram, enhancing face classification models. Key highlights include:

• Utilization of various environment images, including stage setups.
• Inclusion of ‘hard case’ images like faces with masks or hats.
• Extensive tests with state-of-the-art models demonstrating effectiveness.

This dataset not only improves the reliability of facial recognition systems but also enables deeper studies into the nuances of facial attributes in differing conditions. The accessibility of KoIn and its potential applications provide an invaluable resource for further AI-driven research in facial recognition technologies.