A master thesis from Aalborg University
R-FCN Object Detection Ensemble based on Object Resolution and Image Quality
Author(s)
Term
4. term
Publication year
2017
Submitted on
2017-06-08
Abstract
Object detection can be difficult due to challenges such as variations in objects both inter- and intra-class. Additionally, variations can also be present between images. Based on this, research was conducted into creating an ensemble of Region-based Fully Convolutional Networks (R-FCN) object detectors. Ensemble methods explored were firstly data sampling and selection and secondly combination strategies. Data sampling and selection aimed to create different subsets of data with lowered variance with respect to object size and image quality such that expert R-FCN ensemble members could be trained. Two combination strategies were explored for combining the individual member detections into an ensemble result. R-FCNs were trained and tested on the PASCAL VOC benchmark object detection dataset. Results proved positive with an increase in AP when ensemble members were combined appropriately. The method shows potential and other object or image variations could be sampled to see if a more robust ensemble could be made.
Documents
Colophon: This page is part of the AAU Student Projects portal, which is run by Aalborg University. Here, you can find and download publicly available bachelor's theses and master's projects from across the university dating from 2008 onwards. Student projects from before 2008 are available in printed form at Aalborg University Library.
If you have any questions about AAU Student Projects or the research registration, dissemination and analysis at Aalborg University, please feel free to contact the VBN team. You can also find more information in the AAU Student Projects FAQs.