Johanna Wald

Papers

Total Citations

H-Index

About

Johanna Wald is a researcher advancing the frontier of 3D computer vision, with a focus on enabling robust spatial understanding in dynamic, real-world environments. Her work centers on camera and object re-localization—the challenge of accurately determining position and orientation within a scene that has changed over time. In her highly influential 2019 paper, "RIO: 3D Object Instance Re-Localization in Changing Indoor Environments" (9 citations), she introduced a novel task: given objects in one RGB-D scan, estimating their 6DoF poses in a later scan of the same space. This work addresses a critical gap in robotics and augmented reality, where environments are rarely static. Building on this, her 2020 study, "Beyond Controlled Environments: 3D Camera Re-localization in Changing Indoor Scenes" (3 citations), pushes further into the complexities of real-world clutter and lighting variations. Wald’s contributions are foundational for systems that must operate reliably over time, with her RIO benchmark becoming a key reference for the field. Her research is essential reading for anyone working on long-term visual localization or object persistence in AI.