Camouflaged Instance Segmentation In-The-Wild: Dataset, Method, and Benchmark Suite

Abstract

This paper pushes the envelope on camouflaged regions to decompose them into meaningful components, namely, camouflaged instances. To promote the new task of camouflaged instance segmentation, we introduce a new large-scale dataset, namely CAMO++, by extending our preliminary CAMO dataset (camouflaged object segmentation) in terms of quantity and diversity. The new dataset substantially increases the number of images with hierarchical pixel-wise ground-truths. We also provide a benchmark suite for the task of camouflaged instance segmentation. In particular, we conduct extensive evaluation of state-of-the-art instance segmentation detectors on our newly constructed CAMO++ dataset in various scenarios. The dataset, evaluation suite, and benchmark will be publicly available at our project page. 1

GANmouflage: 3D Object Nondetection with Texture Fields

“…Camouflaged object detection. Recent work has sought to detect camouflaged objects using object detectors [15,29,54] and motion cues [8,28]. The focus of our work is generating camouflaged objects, rather than detecting them.…”

OpenForensics: Large-Scale Challenging Dataset For Multi-Face Forgery Detection And Segmentation In-The-Wild

“…Year Object Type #Annotated Images Ground-Truth Type COCO [48] 2014 General object 200,000 Coarse mask CityScapes [11] 2016 Road object 25,000 Coarse&Fine mask WiderFace [77] 2016 Human face 32,200 Bounding box SESIV [37] 2019 Salient object 5,700 Fine mask ADV [38] 2020 Accident object 10,000 Fine mask CAMO++ [36] 2021 Camouflaged object 5,500 Fine mask OpenForensics 2021 Forged face 115,325 Fine mask identities without repeatedly training the AEs. We achieve this by transforming GAN-based high-quality synthesized faces into original poses.…”