Frederik Deroo

Papers

Total Citations

H-Index

About

Frederik Deroo is a researcher advancing the frontier of environment perception for autonomous systems, with a primary focus on semantic occupancy mapping and probabilistic inference. His most-cited work, "Particle-Based Dynamic Semantic Occupancy Mapping Using Bayesian Generalized Kernel Inference" (2024), introduces a novel framework that represents the environment not as a static grid, but as a dynamic, particle-based model capable of capturing both geometric structure and semantic class—such as distinguishing between roads, pedestrians, and vehicles. This approach leverages Bayesian generalized kernel inference to fuse sparse sensor data into a dense, probabilistic map, enabling safer navigation for autonomous vehicles and mobile robots even in highly dynamic settings. Though early in its citation lifecycle, the paper has already garnered attention for its theoretical rigor and practical relevance. Deroo’s contributions are particularly notable for bridging the gap between traditional occupancy grids and modern, learning-based semantic understanding, offering a scalable solution that is both mathematically principled and computationally efficient. His work stands as a key reference for researchers tackling the challenge of robust, real-time environment modeling in intelligent transportation systems.