A Practical Annotation Framework for Short-Form Video Subtitle Translation

This working paper presents a compact annotation and review framework for subtitle translation in short-form video. It focuses on public-facing instructional, educational, and explanatory clips where translators must preserve meaning under strong timing and line-length constraints. The document includes a review rubric, a scene-based multilingual sample set, and a provenance model designed for repository sharing without redistributing third-party audiovisual material. All subtitle examples in the paper were newly authored for this release. No third-party videos, scraped captions, platform transcripts, or copyrighted subtitle files are redistributed in the document. The release is intended as a methodological and teaching resource for subtitle review, translator training, and future annotation work. Project context and implementation source: AI Translate Video. The linked project is relevant here as implementation context for subtitle translation, multilingual video localization, and browser-based review workflows of the general kind discussed in this paper.