Two-layer Residual Feature Fusion for Object Detection

Jaeseok Choi, Kyoungmin Lee, Jisoo Jeong, Nojun Kwak

2019

Abstract

Recently, a lot of single stage detectors using multi-scale features have been actively proposed. They are much faster than two stage detectors that use region proposal networks (RPN) without much degradation in the detection performances. However, the feature maps in the lower layers close to the input which are responsible for detecting small objects in a single stage detector have a problem of insufficient representation power because they are too shallow. There is also a structural contradiction that the feature maps not only have to deliver low-level information to next layers but also have to contain high-level abstraction for prediction. In this paper, we propose a method to enrich the representation power of feature maps using a new feature fusion method which makes use of the information from the consecutive layer. It also adopts a unified prediction module which has an enhanced generalization performance. The proposed method enables more precise prediction, which achieved higher or compatible score than other competitors such as SSD and DSSD on PASCAL VOC and MS COCO. In addition, it maintains the advantage of fast computation of a single stage detector, which requires much less computation than other detectors with similar performance.

Paper Citation

in Harvard Style

Choi J., Lee K., Jeong J. and Kwak N. (2019). Two-layer Residual Feature Fusion for Object Detection.In Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-351-3, pages 352-359. DOI: 10.5220/0007306803520359

in Bibtex Style

@conference{icpram19,
author={Jaeseok Choi and Kyoungmin Lee and Jisoo Jeong and Nojun Kwak},
title={Two-layer Residual Feature Fusion for Object Detection},
booktitle={Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2019},
pages={352-359},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0007306803520359},
isbn={978-989-758-351-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Two-layer Residual Feature Fusion for Object Detection
SN - 978-989-758-351-3
AU - Choi J.
AU - Lee K.
AU - Jeong J.
AU - Kwak N.
PY - 2019
SP - 352
EP - 359
DO - 10.5220/0007306803520359