Osman Tursun
Researcher at QUT SAIVT Lab · “Coding instead of begging”
I work on computer vision and machine learning — including scene text removal and saliency — and build open-source tools for the Uyghur language, from script conversion to text-to-speech.
Selected projects
- Uyghur Multi-Script Converter Converts between Uyghur Arabic, Latin, Cyrillic, Common Turkic and Yengi scripts.
- MTRNet++ One-stage mask-based scene text eraser.
- SESS Saliency Enhancing with Scaling and Sliding.
- METU Trademark Dataset Resources for logo and trademark retrieval research.
- Uyghur Text-to-Speech Simple Uyghur TTS.
- Fonts for Uyghur Arabic Script A collection of fonts for the Uyghur Arabic script.