IIT People Search

Research center

I am a PostDoc researcher in the PAVIS group at the Istituto Italiano di Tecnologia (IIT).

My research is focused on the problem of object-based visual localization. I am interest in how localization and movement through an environment could be approached in terms of the objects present in an environment, rather than a full reconstruction of the scene. Towards this purpose, I am currently focusing on subgraph-matching techniques to compare the estimated locations of an observed subset of object in a camera-centric reference system against the known locations of objects in an environment expressed in a GPS-aligned reference system. Future works will focus on object-based scene reconstruction, and in applying the supgraph-matching techniques to other computer vision problems.

Before starting to work at IIT I obtained a PhD degree in Electronic Engineering from the University of Surrey (UK), working on the topics of human pose estimation, structure from motion and implicit differentiation for nested optimization problems; this work resulted in my  PhD thesis 'Continuous Learning of Inverse Problems with Applications to Structure from Motion', written under the supervision of Dr. Chris Russell and Prof. Adrian Hilton. My prior academic career includes a B.Sc. in Physics from the University of Trieste (IT) and a M.Sc. in Condensed Matter Physics from the University of Trieste. The latter resulted in the master thesis "Simulation of the effects of a habitat with non-constant curvature and variable complexity on the lattice structure of the grid cells and on the mechanisms involved in spatial representation in the hippocampus of a rodent", written in collaboration with the Limbo group from the International School for Advanced Studies (SISSA), under the supervision of Prof. Alessandro Treves.

All Publications
Ahmad J., Taiana M., Toso M., James S., Del Bue A.
Multi-view 3D Objects Localization from Street-level Scenes
International Conference on Image Analysis and Processing
Conference Paper Conference
Ahmad J., Toso M., Taiana M., James S., Del Bue A.
Multi-view 3D Objects Localization from Street-Level Scenes
Lecture Notes in Computer Science, vol. 13232 LNCS, pp. 89-101
Conference Paper Book Series
Taiana M., Toso M., James S., Del Bue A.
PoserNet: Refining Relative Camera Poses Exploiting Object Detections
European Conference on Computer Vision
Article Conference
Taiana M., Toso M., Mehdi M. S., James S., Del Bue A.
The MEMEX project: MEMories and EXperiences for inclusive digital storytelling
Workshop AI Responsabile ed Affidabile - Ital-IA - Secondo Convegno Nazionale CINI sull'Intelligenza Artificiale
Conference Paper Conference
Tome D., Toso M., Agapito L., Russell C.
Rethinking pose in 3D: Multi-stage refinement and recovery for markerless motion capture
Proceedings - 2018 International Conference on 3D Vision, 3DV 2018, pp. 474-483