Camouflaged Object Segmentation

CAMO Dataset V-1.0

Camouflaged Object (CAMO) dataset specifically designed for the task of camouflaged object segmentation. We focus on two categories, i.e., naturally camouflaged objects and artificially camouflaged objects, which usually correspond to animals and humans in the real world, respectively. The CAMO dataset consists of 1250 images (1000 images for the training set and 250 images for the testing set).

CAMO-COCO Dataset V-1.0

CAMO-COCO dataset consisting of camouflaged object images and non-camouflaged object images. We use the entire images in the CAMO dataset for the camouflaged object images and collected additional 1250 images from the MS-COCO dataset for the non-camouflaged object images. The CAMO-COCO dataset consists of 2500 images (2000 images for the training set and 500 images for the testing set).

ANet

Downloads

Datasets: [CAMO-V.1.0] [CAMO-COCO-V.1.0] [Evaluation Code]

Pre-Computed Results: [ANet]

If you use our datasets or results, please cite the following publications:

@article{ltnghia-CVIU2019,   Title          = {Anabranch Network for Camouflaged Object Segmentation},   Author         = {Trung-Nghia Le and Tam V. Nguyen and Zhongliang Nie and Minh-Triet Tran and Akihiro Sugimoto},   Journal        = {Journal of Computer Vision and Image Understanding}   Year           = {2019}   Volume         = {184},    Pages          = {45-56},  }

Publications

Trung-Nghia Le, Tam V. Nguyen, Zhongliang Nie, Minh-Triet Tran, Akihiro Sugimoto, "Anabranch Network for Camouflaged Object Segmentation", Journal of Computer Vision and Image Understanding (CVIU), 2019. [PDF]