PhD Researcher in Computer Vision at Free University of Bozen-Bolzano 🎓
Specializing in Multimodal Video Understanding and Vision-Language Models for sports performance analysis 🏀⚽

🔬 Research Focus: Efficient architectures for action quality assessment, skill evaluation, and AI-driven feedback generation
✍️ Technical Writer @ Towards AI • 20+ articles • 25K+ reads
🚀 Building lightweight VLMs with state-of-the-art performance

• SkillFormer – Multi-view action quality assessment (4.5× fewer parameters)
• PATS – Proficiency-aware temporal sampling (+26% performance gains)
• ProfVLM – Lightweight vision-language model for skill assessment (20× fewer parameters)

• PATS: Proficiency-Aware Temporal Sampling for Multi-View Sports Skill Assessment (IEEE STAR 2025)
• SkillFormer: Unified Multi-View Video Understanding for Proficiency Estimation (ICMV 2025)
• Gate-Shift-Pose: Enhancing Action Recognition in Sports with Skeleton Information. (WACVW 2025)