Zenithal isotropic object counting by localization using adversarial training

Counting objects in images is a very time-consuming task for humans that yields to errors caused by repetitiveness and boredom. In this paper, we present a novel object counting method that, unlike most of the recent works that focus on the regression of a density map, performs the counting procedure by localizing each single object. This key difference allows us to provide not only an accurate count but the position of every counted object, information that can be critical in some areas such as precision agriculture. The method is designed in two steps: first, a CNN is in charge of mapping arbitrary objects to blob-like structures. Then, using a Laplacian of Gaussian (LoG) filter, we are able to gather the position of all detected objects. We also propose a semi-adversarial training procedure that, combined with the former design, improves the result by a large margin. After evaluating the method on two public benchmarks of isometric objects, we stay on par with the state of the art while being able to provide extra position information. (C) 2021 Elsevier Ltd. All rights reserved.

Automated summary

This paper presents a novel object counting method that provides accurate counts and object position information by localizing each object. The method first maps objects to blob-like structures using CNN, then gathers object positions using a LoG filter, and improves results significantly through a semi-adversarial training procedure. The method performs on par with the state of the art while offering additional position information.