Benutzer: Gast  Login
Titel:

3D Object Instance Recognition and Pose Estimation Using Triplet Loss with Dynamic Margin

Dokumenttyp:
Konferenzbeitrag
Autor(en):
Zakharov, S.; Kehl, W.; Planche, B.; Hutter, A.; Ilic, S.
Abstract:
In this paper, we address the problem of 3D object instance recognition and pose estimation of localized objects in cluttered environments using convolutional neural networks. Inspired by the descriptor learning approach of Wohlhart et al. i̧tewohlhart2015learning, we propose a method that introduces the dynamic margin in the manifold learning triplet loss function. Such a loss function is designed to map images of different objects under different poses to a lower-dimensional, similarity-prese...     »
Stichworte:
CAMP,CAMPComputerVision,ComputerVision,IROS,3DObjectInstanceRecognition,3DPoseEstimation
Kongress- / Buchtitel:
Proceedings of the International Conference on Intelligent Robots and Systems
Jahr:
2017
Hinweise:
Open attachment browser
 BibTeX