Publication
Depth-aware convolutional neural networks for accurate 3D pose estimation in RGB-D images
Conference Article
Conference
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Edition
2017
Pages
5777-5783
Doc link
http://dx.doi.org/10.1109/IROS.2017.8206469
File
Authors
-
Porzi, Lorenzo
-
Peñate Sánchez, Adrián
-
Ricci, Elisa
-
Moreno Noguer, Francesc
Projects associated
AEROARMS: AErial RObotics System integrating multiple ARMS and advanced manipulation capabilities for inspection and maintenance
RobInstruct: Instructing robots using natural communication skills
I-DRESS: Assistive interactive robotic system for support in dressing
MdM: Unit of Excellence María de Maeztu
Abstract
Most recent approaches to 3D pose estimation from RGB-D images address the problem in a two-stage pipeline. First, they learn a classifier –typically a random forest– to predict the position of each input pixel on the object surface. These estimates are then used to define an energy function that is minimized w.r.t. the object pose. In this paper, we focus on the first stage of the problem and propose a novel classifier based on a depth-aware Convolutional Neural Network. This classifier is able to learn a scale-adaptive regression model that yields very accurate pixel-level predictions, allowing to finally estimate the pose using a simple RANSAC-based scheme, with no need to optimize complex ad hoc energy functions. Our experiments on publicly available datasets show that our approach achieves remarkable improvements over state-of-the-art methods.
Categories
computer vision.
Author keywords
Deep Learning
Scientific reference
L. Porzi, A. Penate-Sanchez, E. Ricci and F. Moreno-Noguer. Depth-aware convolutional neural networks for accurate 3D pose estimation in RGB-D images, 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2017, Vancouver, Canada, pp. 5777-5783.
Follow us!