Where are my clothes? A multi-level approach for evaluating deep instance segmentation architectures on fashion images
JOUANNEAU, Warren
Lectra
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Université de Bordeaux [UB]
Voir plus >
Lectra
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Université de Bordeaux [UB]
JOUANNEAU, Warren
Lectra
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Université de Bordeaux [UB]
< Réduire
Lectra
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Université de Bordeaux [UB]
Langue
en
Communication dans un congrès
Ce document a été publié dans
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, CVPR Workshop on Computer Vision for Fashion, Art, and Design (CVFAD'21), 2021-06-19, Virtual event. 2021-06-19p. 3951-3955
Résumé en anglais
In this paper we present an extensive evaluation of instance segmentation in the context of images containing clothes. We propose a multi level evaluation that completes the classical overlapping criteria given by IoU. In ...Lire la suite >
In this paper we present an extensive evaluation of instance segmentation in the context of images containing clothes. We propose a multi level evaluation that completes the classical overlapping criteria given by IoU. In particular, we quantify both the contour and color content accuracy of the the predicted segmentation masks. We demonstrate that the proposed evaluation framework is relevant to obtain meaningful insights on models performance through experiments conducted on five state of the art instance segmentation methods.< Réduire
Origine
Importé de halUnités de recherche