Mauro Comi
Papers
2
Total Citations
25
H-Index
2
About
Mauro Comi is a researcher at the forefront of robotic tactile sensing and 3D shape reconstruction, bridging the gap between vision and touch for autonomous systems. His primary research areas include vision-based tactile sensing, sim-to-real transfer for robotic control, and data-driven 3D shape understanding. Comi’s most notable contribution is **TouchSDF** (2024, 21 citations), a pioneering method that applies DeepSDF—a neural implicit representation—to reconstruct 3D object shapes from high-resolution tactile data alone. This work demonstrates how robots can leverage tactile feedback to infer geometry, mimicking human reliance on touch for environmental understanding. In parallel, his paper **“Attention for Robot Touch”** (2023, 4 citations) introduces tactile saliency prediction, a novel attention mechanism that enhances robustness in contact-rich robotic tasks during sim-to-real deployment. By focusing on salient tactile features, this approach improves control in unstructured environments—a critical challenge for real-world robotics. Comi’s work is distinguished by its integration of deep learning with physical sensing, offering scalable solutions for manipulation and exploration. His research not only advances tactile perception but also provides foundational tools for robots to interact safely and intelligently with their surroundings, making him a key voice in embodied AI and robotic dexterity.
Research Focus
Key Achievements
Top Papers
- 1
- 2