Zeno Sambugaro

EVS Embedded Vision Systems · Verona, Italy

At EVS, I build in-cabin occupant monitoring systems for autonomous driving, covering the full pipeline from data to a deployed model running on TensorFlow Lite and ONNX. Along the way I've built real-time object detectors (YOLO, DETR) tuned for embedded hardware, and designed a 3D human pose and shape estimation system (SMPL) that relies on precise multi-view geometry and camera calibration to read movement in 3D space. Most recently, I led the core algorithms behind a 3D human-mesh reconstruction system that we demoed live at CES 2026 in Las Vegas.

Simon Fraser University · Vancouver, Canada

For six months, I worked in the GrUVi Lab at Simon Fraser University under Prof. Andrea Tagliasacchi, developing a new method for fast, efficient 3D rendering and representation built on Instant-NGP. That work was later published at ECCV 2024.

University of Trento · Trento, Italy

Thesis: “The Role of Interpretable Neural Architectures: From Image Classification to Neural Fields.” Advised by Prof. Nicola Conci.

University of Trento · Trento, Italy

Final grade: 110/110 cum laude. Thesis: “Out-of-Distribution Detection in Videos Using a Union of 1-Dimensional Subspaces.”

University of Trento · Trento, Italy