I build and evaluate machine-learning systems across Computer Vision, NLP, and Explainable AI, with a focus on reproducible experiments and well-structured research software.
• Education: M.Sc. student at Politecnico di Torino, specializing in Artificial Intelligence and Data Analytics.
• Background: B.Sc. in Applied Computer Science from the University of Urbino, with research experience in Human Activity Recognition for wearable devices.
• Interests: Deep Learning, Computer Vision, NLP, Explainable AI, anomaly detection, and reliable ML evaluation.
• Engineering approach: I enjoy turning experimental ideas into configurable, documented, and reproducible pipelines.
| Project | Area | Description |
|---|---|---|
| MAAS-Road-Scenes | Computer Vision · OOD Detection | Evaluation and analysis framework for road-scene anomaly segmentation, comparing ERFNet and EoMT across multiple benchmarks and fine-grained failure modes. |
| CLARITY | NLP · LLMs | SemEval 2026 research project for detecting evasive political answers using transformer encoders, prompted LLMs, and parameter-efficient fine-tuning. |
| EQE | Explainable AI | Experimental framework for evaluating whether compact, human-readable explanations remain faithful to black-box model predictions. |
| InformaticaUniurb | Open Source · Education | Collaborative collection of Computer Science notes and learning resources created for university students. |
| Minerva | Computer Vision · ML Engineering | Work in Progress: modular backend for multi-attribute clothing classification using a shared ResNet-50 backbone and multiple prediction heads. |
• Machine Learning: PyTorch, Hugging Face Transformers, scikit-learn, XGBoost, SHAP, LIME.
• Languages: Python, Java, C, Rust.
• Engineering: Docker, Linux, Git, GitHub, uv, OmegaConf.
• Data and Web: MySQL, REST APIs, HTML, CSS.
• Developing Minerva, a modular Computer Vision backend for clothing-attribute recognition.
• Exploring robust evaluation methods for out-of-distribution detection and Explainable AI.
• Improving the reproducibility, documentation, and engineering quality of research-oriented ML projects.
View contribution metrics
