Standard

Divide and Count : Generic Object Counting by Image Divisions. / Stahl, Tobias; Pintea, Silvia L.; Van Gemert, Jan C.

In: IEEE Transactions on Image Processing, Vol. 28, No. 2, 8488575, 2019, p. 1035-1044.

Research output: Contribution to journalArticleScientificpeer-review

Harvard

Stahl, T, Pintea, SL & Van Gemert, JC 2019, 'Divide and Count: Generic Object Counting by Image Divisions' IEEE Transactions on Image Processing, vol. 28, no. 2, 8488575, pp. 1035-1044. https://doi.org/10.1109/TIP.2018.2875353

APA

Vancouver

Author

Stahl, Tobias ; Pintea, Silvia L. ; Van Gemert, Jan C. / Divide and Count : Generic Object Counting by Image Divisions. In: IEEE Transactions on Image Processing. 2019 ; Vol. 28, No. 2. pp. 1035-1044.

BibTeX

@article{fe82df6596eb4461b03697f63aecea21,
title = "Divide and Count: Generic Object Counting by Image Divisions",
abstract = "We propose a general object counting method that does not use any prior category information. We learn from local image divisions to predict global image-level counts without using any form of local annotations. Our method separates the input image into a set of image divisions - each fully covering the image. Each image division is composed of a set of region proposals or uniform grid cells. Our approach learns in an end-to-end deep learning architecture to predict global image-level counts from local image divisions. The method incorporates a counting layer which predicts object counts in the complete image, by enforcing consistency in counts when dealing with overlapping image regions. Our counting layer is based on the inclusion-exclusion principle from set theory. We analyze the individual building blocks of our proposed approach on Pascal-VOC2007 and evaluate our method on the MS-COCO large scale generic object data set as well as on three class-specific counting data sets: UCSD pedestrian data set, and CARPK, and PUCPR+ car data sets.",
keywords = "counting with region proposals, fully convolutional networks, Generic-class object counting, inclusion-exclusion principle, regression",
author = "Tobias Stahl and Pintea, {Silvia L.} and {Van Gemert}, {Jan C.}",
note = "Accepted Author Manuscript",
year = "2019",
doi = "10.1109/TIP.2018.2875353",
language = "English",
volume = "28",
pages = "1035--1044",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

RIS

TY - JOUR

T1 - Divide and Count

T2 - IEEE Transactions on Image Processing

AU - Stahl, Tobias

AU - Pintea, Silvia L.

AU - Van Gemert, Jan C.

N1 - Accepted Author Manuscript

PY - 2019

Y1 - 2019

N2 - We propose a general object counting method that does not use any prior category information. We learn from local image divisions to predict global image-level counts without using any form of local annotations. Our method separates the input image into a set of image divisions - each fully covering the image. Each image division is composed of a set of region proposals or uniform grid cells. Our approach learns in an end-to-end deep learning architecture to predict global image-level counts from local image divisions. The method incorporates a counting layer which predicts object counts in the complete image, by enforcing consistency in counts when dealing with overlapping image regions. Our counting layer is based on the inclusion-exclusion principle from set theory. We analyze the individual building blocks of our proposed approach on Pascal-VOC2007 and evaluate our method on the MS-COCO large scale generic object data set as well as on three class-specific counting data sets: UCSD pedestrian data set, and CARPK, and PUCPR+ car data sets.

AB - We propose a general object counting method that does not use any prior category information. We learn from local image divisions to predict global image-level counts without using any form of local annotations. Our method separates the input image into a set of image divisions - each fully covering the image. Each image division is composed of a set of region proposals or uniform grid cells. Our approach learns in an end-to-end deep learning architecture to predict global image-level counts from local image divisions. The method incorporates a counting layer which predicts object counts in the complete image, by enforcing consistency in counts when dealing with overlapping image regions. Our counting layer is based on the inclusion-exclusion principle from set theory. We analyze the individual building blocks of our proposed approach on Pascal-VOC2007 and evaluate our method on the MS-COCO large scale generic object data set as well as on three class-specific counting data sets: UCSD pedestrian data set, and CARPK, and PUCPR+ car data sets.

KW - counting with region proposals

KW - fully convolutional networks

KW - Generic-class object counting

KW - inclusion-exclusion principle

KW - regression

UR - http://www.scopus.com/inward/record.url?scp=85054641664&partnerID=8YFLogxK

U2 - 10.1109/TIP.2018.2875353

DO - 10.1109/TIP.2018.2875353

M3 - Article

VL - 28

SP - 1035

EP - 1044

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 2

M1 - 8488575

ER -

ID: 47407265