DeepTrace — South Asia Deepfake Detection
Annotation guidelines and benchmarks for AI-generated visual content led by Prof. Dilrukshi Gamage
A volunteer research contribution to DeepTrace, a project on deepfake detection in South Asian contexts (November 2025 – January 2026).
I worked togther with Team establish annotation guidelines and benchmarks for AI-generated visual content and deepfake detection, and contributed to the design of the dataset-collection interface. The work addresses a persistent gap in detection research: most benchmarks under-represent South Asian faces, languages, and visual contexts.
Role: Research assistant — annotation protocol, benchmarking, dataset UI/UX. Project: South Asia Deepfake Detection Research Mentor:: [Dilrukshi Gamage]https://www.dilrukshigamage.org/)