Tomasz Malisiewicz

Papers

Total Citations

180

H-Index

About

Tomasz Malisiewicz is a leading computer vision researcher whose work bridges the gap between classical geometric methods and modern deep learning, with a focus on 3D perception, pose estimation, and visual localization. His most impactful contribution is **OrienterNet** (2023, 78 citations), a pioneering deep neural network that enables visual localization using simple 2D public maps instead of expensive 3D point clouds—a breakthrough that makes spatial orientation accessible for everyday applications. Malisiewicz also developed **Deep ChArUco** (2019, 80 citations), which extends traditional ChArUco fiducial markers to perform robustly in low-light conditions, solving a critical failure point in robotics and augmented reality pipelines. His earlier work on **Deep Cuboid Detection** (2016, 18 citations) introduced an end-to-end deep learning approach for localizing 3D box-like objects from single RGB images, moving beyond 2D bounding boxes. With over 180 total citations, Malisiewicz’s research consistently pushes the boundaries of practical 3D vision, making complex geometric problems solvable with consumer-grade hardware. His work is widely used in OpenCV and has direct applications in autonomous navigation, AR, and industrial robotics.