Bernard Lange

Papers

Total Citations

H-Index

About

Bernard Lange is a researcher at the forefront of autonomous vehicle perception and prediction, with a focus on developing scalable, self-supervised frameworks for dynamic environment modeling. His most cited work, "LOPR: Latent Occupancy PRediction using Generative Models" (2022), introduces a novel approach that leverages LiDAR-generated occupancy grid maps (L-OGMs) to predict future scene states without the need for costly manual annotations. By employing generative models in a latent space, Lange’s method enables robust, joint scene predictions in bird’s-eye view, directly addressing a critical bottleneck in autonomous navigation: the reliance on labeled data. This contribution has garnered early recognition with 4 citations, signaling its growing influence in the field. Lange’s research bridges the gap between perception and planning, offering a path toward safer, more adaptable autonomous systems. His work is particularly notable for its emphasis on unsupervised learning, which reduces human effort while improving generalization across diverse driving environments. For students and researchers, Lange exemplifies how generative AI can be harnessed to solve real-world robotics challenges, making his profile a compelling study in innovation at the intersection of computer vision and autonomous systems.