Sciences Informatiques : Les résultats de la recherche académique financée par l'IMRA presentés à l'ICIAP 2019
Holistic trajectories shown on the reference frame. Past: full squares. Present: full circle. Future: empty squares.
Last September at the ICIAP'19 conference at Trento, Italy, our collaborators from the MICC group at the University of Florence, Italy presented some results of their work performed during their research on multiple future prediction funded by IMRA Europe S.A.S.
This work is entitled “Vehicle Trajectories from Unlabeled Data Through Iterative Plane Registration” and presents a novel approach for extracting the full trajectories of moving objects from egocentric views, to be used as ground truth for the supervised learning of their multiple future prediction network.
Congratulations to all the team of MICC for this great academic achievement!
Article and related material: here (ICIAP 2019 at Springer Link)