Depth Value Pre-Processing for Accurate Transfer Learning based RGB-D Object Recognition

Andreas Aakerberg, Kamal Nasrollahi, Christoffer B. Rasmussen, Thomas B. Moeslund

Abstract

Object recognition is one of the important tasks in computer vision which has found enormous applications.Depth modality is proven to provide supplementary information to the common RGB modality for objectrecognition. In this paper, we propose methods to improve the recognition performance of an existing deeplearning based RGB-D object recognition model, namely the FusionNet proposed by Eitel et al. First, we showthat encoding the depth values as colorized surface normals is beneficial, when the model is initialized withweights learned from training on ImageNet data. Additionally, we show that the RGB stream of the FusionNetmodel can benefit from using deeper network architectures, namely the 16-layered VGGNet, in exchange forthe 8-layered CaffeNet. In combination, these changes improves the recognition performance with 2.2% incomparison to the original FusionNet, when evaluating on the Washington RGB-D Object Dataset.

Download


Paper Citation


in Harvard Style

Aakerberg A., Nasrollahi K., Rasmussen C. and Moeslund T. (2017). Depth Value Pre-Processing for Accurate Transfer Learning based RGB-D Object Recognition.In Proceedings of the 9th International Joint Conference on Computational Intelligence - Volume 1: IJCCI, ISBN 978-989-758-274-5, pages 121-128. DOI: 10.5220/0006511501210128


in Bibtex Style

@conference{ijcci17,
author={Andreas Aakerberg and Kamal Nasrollahi and Christoffer B. Rasmussen and Thomas B. Moeslund},
title={Depth Value Pre-Processing for Accurate Transfer Learning based RGB-D Object Recognition},
booktitle={Proceedings of the 9th International Joint Conference on Computational Intelligence - Volume 1: IJCCI,},
year={2017},
pages={121-128},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006511501210128},
isbn={978-989-758-274-5},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 9th International Joint Conference on Computational Intelligence - Volume 1: IJCCI,
TI - Depth Value Pre-Processing for Accurate Transfer Learning based RGB-D Object Recognition
SN - 978-989-758-274-5
AU - Aakerberg A.
AU - Nasrollahi K.
AU - Rasmussen C.
AU - Moeslund T.
PY - 2017
SP - 121
EP - 128
DO - 10.5220/0006511501210128