I'm a Ph.D. candidate at Técnico Lisboa (University of Lisbon), specializing in Computer Vision.

My research focuses on 3D Reconstruction, SLAM, and Robust Estimation. I am dedicated to designing and implementing high-performance algorithms that enable reliable, real-time perception and localization for real-world applications.

(All selected publications include open-source code/repositories)

Revisiting Monocular SLAM with Spatio-Temporal Scene Modeling.
Valter Piedade, Lalit Manam, Masashi Yamazaki, Pedro Miraldo.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.
[Project · Paper · Repository TBA]

SAC-GNC: SAmple Consensus for adaptive Graduated Non-Convexity. Highlight paper (2.3%)
Valter Piedade, Chitturi Sidhartha, José António Gaspar, Venu Madhav Govindu, Pedro Miraldo.
IEEE/CVF International Conference on Computer Vision (ICCV), 2025.
[Project · Paper · Code]

BANSAC: A Dynamic BAyesian Network for Adaptive SAmple Consensus.
Valter Piedade, Pedro Miraldo.
IEEE/CVF International Conference on Computer Vision (ICCV), 2023.
[Project · Paper · Code]