Marco Rosano

About

Marco Rosano is an emerging researcher specializing in embodied visual navigation and robot perception, with a particular focus on bridging the gap between simulation-based training and real-world deployment of autonomous systems. His work addresses one of the most pressing challenges in robotics and artificial intelligence: enabling deep learning models to generalize effectively when transferred from simulated environments to complex, unstructured real-world settings. Rosano's most notable contribution, "Image-based Navigation in Real-World Environments via Multiple Mid-level Representations" (2023), introduces fusion models and benchmark frameworks that leverage mid-level visual representations to improve navigation robustness, earning 5 citations since its publication. His earlier work, "On Embodied Visual Navigation in Real Environments Through Habitat" (2021), further explores how reinforcement learning policies trained within simulation platforms can be adapted for physical robotic deployment — a critical step toward practical autonomous navigation systems. Through his research, Rosano contributes meaningfully to the fields of computer vision, reinforcement learning, and mobile robotics. His emphasis on efficient evaluation methodologies and real-world benchmarking makes his work particularly valuable for researchers seeking to deploy intelligent navigation systems beyond the laboratory.

Research Focus

Key Achievements

Top Papers

1. 1
   Image-based Navigation in Real-World Environments via Multiple Mid-level Representations: Fusion Models, Benchmark and Efficient Evaluation

   5 citations · 2023
2. 2
   On Embodied Visual Navigation in Real Environments Through Habitat

   2 citations · 2021

Key Collaborators

Contact & Links