| Title |
Camouflaged object detection using convolutional neural network |
| Authors |
Mickevičius, Karolis ; Lipnickas, Arūnas |
| Full Text |
|
| Is Part of |
CEUR Workshop proceedings: IVUS 2025: proceedings of the 30th international conference on information society and university studies (IVUS 2025), Kaunas, Lithuania, 15 May 2025 / edited by: I. Veitaitė, A. Lopata, T. Krilavičius, M. Woźniak.. Aachen : CEUR-WS. 2026, vol. 4213, p. 40-49.. ISSN 1613-0073 |
| Keywords [eng] |
Neural network ; convolutional neural network ; U-Net ; Camouflaged Object Detection ; image segmentation |
| Abstract [eng] |
Camouflaged object detection assists in the detection of camouflaged animals, military personnel and equipment. The usage of such algorithms extends to finding specified cells in biology and detection of surface defects in industry. This study researches the task of detecting camouflaged objects using a convolutional neural network. A U-Net network is selected and adapted to receive a dataset of camouflaged objects and is tested to be able to process it. Different U-Net modifications, loss functions, photo sizes, colour channels and conversion spaces for camouflaged object detection are studied. The results of camouflaged object detection are shown as Dice, IoU and AUC metrics. The results are compared and it is found accurately did the U-Net model predict the location of camouflaged objects. |
| Published |
Aachen : CEUR-WS |
| Type |
Conference paper |
| Language |
English |
| Publication date |
2026 |
| CC license |
|
