Drawlyzer: ai-based drawing assessment for art admissions

The Drawlyzer App is an innovative mobile application developed to streamline and standardize the drawing assessment process for prospective students applying to enroll in the Faculty of Art and Design (FSSR) at Universiti Teknologi MARA (UiTM). Traditional manual evaluations often suffer from subjectivity and inconsistency, making talent selection less reliable. Drawlyzer addresses this issue by integrating artificial intelligence to analyze and score hand-drawn artworks based on direct comparison with a reference image. Built using the Expo mobile development framework and powered by a large language model from Anthropic’s Claude 3.7 Sonnet, the app evaluates submitted drawings using visual criteria such as proportion, detail, and technique. Instead of letter grades, the app generates a percentage score along with individual scores for each criterion and provides automated feedback on elements like composition, balance, and line quality. All data is securely stored in the cloud via Supabase for centralized access and reporting. The app prototype has shown promising accuracy in aligning with expert judgment, enhancing fairness and consistency in the admission process. Drawlyzer holds strong commercialization potential across art institutions, digital portfolio reviews, and online art competitions. It supports a transparent, scalable, and efficient approach to talent evaluation while aligning with Sustainable Development Goals related to quality education and innovation.