Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/134826
Citations | ||
Scopus | Web of Science® | Altmetric |
---|---|---|
?
|
?
|
Type: | Conference paper |
Title: | Dropout Sampling for Robust Object Detection in Open-Set Conditions |
Author: | Miller, D. Nicholson, L. Dayoub, F. Sunderhauf, N. |
Citation: | IEEE International Conference on Robotics and Automation, 2018, pp.3243-3249 |
Publisher: | IEEE |
Issue Date: | 2018 |
Series/Report no.: | IEEE International Conference on Robotics and Automation ICRA |
ISBN: | 9781538630815 |
ISSN: | 1050-4729 2577-087X |
Conference Name: | IEEE International Conference on Robotics and Automation (ICRA) (21 May 2018 - 25 May 2018 : Brisbane, Australia) |
Statement of Responsibility: | Dimity Miller, Lachlan Nicholson, Feras Dayoub, Niko Sunderhauf |
Abstract: | Dropout Variational Inference, or Dropout Sampling, has been recently proposed as an approximation technique for Bayesian Deep Learning and evaluated for image classification and regression tasks. This paper investigates the utility of Dropout Sampling for object detection for the first time. We demonstrate how label uncertainty can be extracted from a state-of-the-art object detection system via Dropout Sampling. We evaluate this approach on a large synthetic dataset of 30,000 images, and a real-world dataset captured by a mobile robot in a versatile campus environment. We show that this uncertainty can be utilized to increase object detection performance under the open-set conditions that are typically encountered in robotic vision. A Dropout Sampling network is shown to achieve a 12.3% increase in recall (for the same precision score as a standard network) and a 15.1% increase in precision (for the same recall score as the standard network). |
Rights: | ©2018 IEEE |
DOI: | 10.1109/ICRA.2018.8460700 |
Grant ID: | http://purl.org/au-research/grants/arc/CE140100016 |
Published version: | https://ieeexplore.ieee.org/Xplore/home.jsp |
Appears in Collections: | Computer Science publications |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.