Deep Learning for Vehicle Detection in Aerial Images

M.Y. Yang; Wentong Liao; Xinbo Li; Bodo Rosenhahn

doi:10.1109/ICIP.2018.8451454

Deep Learning for Vehicle Detection in Aerial Images

M.Y. Yang, Wentong Liao, Xinbo Li, Bodo Rosenhahn

Research output: Chapter in Book/Report/Conference proceeding › Chapter › Academic › peer-review

30 Citations (Scopus)

256 Downloads (Pure)

Abstract

The detection of vehicles in aerial images is widely applied in many domains. In this paper, we propose a novel double focal loss convolutional neural network framework (DFL-CNN). In the proposed framework, the skip connection is used in the CNN structure to enhance the feature learning. Also, the focal loss function is used to substitute for conventional cross entropy loss function in both of the region proposed network and the final classifier. We further introduce the first large-scale vehicle detection dataset ITCVD with ground truth annotations for all the vehicles in the scene. The experimental results show that our DFL-CNN outperforms the baselines on vehicle detection.

Original language	English
Title of host publication	2018 25th IEEE International Conference on Image Processing (ICIP)
Subtitle of host publication	7-10 October 2018, Athens, Greece
Publisher	IEEE
Pages	3079-3083
ISBN (Electronic)	978-1-4799-7061-2
DOIs	https://doi.org/10.1109/ICIP.2018.8451454
Publication status	Published - Oct 2018
Event	25th IEEE International Conference on Image Processing, ICIP 2018 - Athens, Greece Duration: 7 Oct 2018 → 10 Oct 2018 Conference number: 25 https://2018.ieeeicip.org/

Conference

Conference	25th IEEE International Conference on Image Processing, ICIP 2018
Abbreviated title	ICIP
Country/Territory	Greece
City	Athens
Period	7/10/18 → 10/10/18
Internet address	https://2018.ieeeicip.org/

Keywords

2021 OA procedure

Access to Document

10.1109/ICIP.2018.8451454

yang_deeFinal published version, 1.89 MBLicence: Taverne

Cite this

@inbook{6069db85ddfb4e4fa1bfe722f625b3fc,

title = "Deep Learning for Vehicle Detection in Aerial Images",

abstract = "The detection of vehicles in aerial images is widely applied in many domains. In this paper, we propose a novel double focal loss convolutional neural network framework (DFL-CNN). In the proposed framework, the skip connection is used in the CNN structure to enhance the feature learning. Also, the focal loss function is used to substitute for conventional cross entropy loss function in both of the region proposed network and the final classifier. We further introduce the first large-scale vehicle detection dataset ITCVD with ground truth annotations for all the vehicles in the scene. The experimental results show that our DFL-CNN outperforms the baselines on vehicle detection.",

keywords = "2021 OA procedure",

author = "M.Y. Yang and Wentong Liao and Xinbo Li and Bodo Rosenhahn",

year = "2018",

month = oct,

doi = "10.1109/ICIP.2018.8451454",

language = "English",

pages = "3079--3083",

booktitle = "2018 25th IEEE International Conference on Image Processing (ICIP)",

publisher = "IEEE",

address = "United States",

note = "25th IEEE International Conference on Image Processing, ICIP 2018, ICIP ; Conference date: 07-10-2018 Through 10-10-2018",

url = "https://2018.ieeeicip.org/",

}

TY - CHAP

T1 - Deep Learning for Vehicle Detection in Aerial Images

AU - Yang, M.Y.

AU - Liao, Wentong

AU - Li, Xinbo

AU - Rosenhahn, Bodo

N1 - Conference code: 25

PY - 2018/10

Y1 - 2018/10

N2 - The detection of vehicles in aerial images is widely applied in many domains. In this paper, we propose a novel double focal loss convolutional neural network framework (DFL-CNN). In the proposed framework, the skip connection is used in the CNN structure to enhance the feature learning. Also, the focal loss function is used to substitute for conventional cross entropy loss function in both of the region proposed network and the final classifier. We further introduce the first large-scale vehicle detection dataset ITCVD with ground truth annotations for all the vehicles in the scene. The experimental results show that our DFL-CNN outperforms the baselines on vehicle detection.

AB - The detection of vehicles in aerial images is widely applied in many domains. In this paper, we propose a novel double focal loss convolutional neural network framework (DFL-CNN). In the proposed framework, the skip connection is used in the CNN structure to enhance the feature learning. Also, the focal loss function is used to substitute for conventional cross entropy loss function in both of the region proposed network and the final classifier. We further introduce the first large-scale vehicle detection dataset ITCVD with ground truth annotations for all the vehicles in the scene. The experimental results show that our DFL-CNN outperforms the baselines on vehicle detection.

KW - 2021 OA procedure

UR - https://ezproxy2.utwente.nl/login?url=https://webapps.itc.utwente.nl/library/2018/chap/yang_dee.pdf

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1109/ICIP.2018.8451454

U2 - 10.1109/ICIP.2018.8451454

DO - 10.1109/ICIP.2018.8451454

M3 - Chapter

SP - 3079

EP - 3083

BT - 2018 25th IEEE International Conference on Image Processing (ICIP)

PB - IEEE

T2 - 25th IEEE International Conference on Image Processing, ICIP 2018

Y2 - 7 October 2018 through 10 October 2018

ER -

Deep Learning for Vehicle Detection in Aerial Images

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this