Image Content Annotation & Moderation for Korean AI Training Project
Worked on a Korean-language AI training project focused on image-based content moderation and computer vision data preparation. The scope of the project involved reviewing and labeling large volumes of user-generated images according to detailed classification guidelines. Tasks included identifying visual attributes, detecting objects and sensitive elements, assigning content categories, and evaluating emotional tone where applicable. Maintained high annotation accuracy by strictly adhering to updated labeling rules and operational policies. Regularly incorporated quality assurance feedback, performed self-checks, and ensured consistency across annotations to support reliable model training outcomes. The project required strong attention to detail, multilingual understanding, and the ability to work efficiently in a remote, production-level environment.