Titre: DRAFT – Distilled Recurrent All-Pairs Field Transforms for Optical Flow
Abstract
The presentation addresses the challenge of utilizing learning-based algorithms for 3D scene reconstruction on resource-constrained end-user devices. Although integrating deep learning methods into the reconstruction pipeline has demonstrated superior performance to classical techniques, the resulting large models could be impractical for resource-limited devices. We propose an efficient solution by introducing a method to compress deep learning models used in 3D reconstruction workflows. Our approach, named DRAFT, employs knowledge distillation (KD), adapted and extended for complex feature and context extraction tasks related to optical flow. New distillation components based on algebraic sign-pattern matrices (SPM) and inertia has been introduced to enhance the KD process.
Intervenant
Yanick Christian TCHENKO, doctorant IBISC, équipe IRA2, supervisé par Hedi TABIA (PR IBISC, équipe IRA2) et Hicham HADJ-ABDELKADER (MCF IBISC, équipe SIAM)
NOTA
Les séminaires IRA2 sont organisés tous les premiers lundi de chaque mois, à l’initiative de la direction de l’équipe IRA2.
- Date: 08/04/2024, 13h30
- Lieu: IBISC, site Pelvoux, salle Ax101, voir le sémininaire sur Zoom: https://univ-evry-fr.zoom.us/j/95109429789?pwd=TEFjUERINWhpdnFhNDRvYmtkbm8yQT09
- Organisation: Hédi TABIA (PR Univ. Évry, IBISC équipe IRA2)
