CoCOD8K
收藏资源简介:
CoCOD8K是由东北石油大学电气信息工程系创建的大型数据集,包含8528张高质量、精心挑选的图像,覆盖5个超类和70个子类。数据集涵盖了自然和人工伪装场景,具有多样化的物体外观和背景,为协同伪装物体检测提供了挑战性的数据基础。数据集的创建旨在推动伪装物体检测领域的发展,解决在多个相关图像中同时检测具有相同属性的伪装物体的难题。
CoCOD8K is a large-scale dataset created by the Department of Electrical and Information Engineering, Northeast Petroleum University. It includes 8528 high-quality, carefully selected images spanning 5 super-categories and 70 sub-categories. The dataset covers both natural and artificial camouflage scenarios, with diverse object appearances and backgrounds, providing a challenging data foundation for collaborative camouflage object detection. The dataset is developed to advance the field of camouflage object detection, addressing the challenge of simultaneously detecting camouflaged objects with the same attributes across multiple relevant images.

- 1Collaborative Camouflaged Object Detection: A Large-Scale Dataset and Benchmark东北石油大学电气信息工程系,中国大庆 · 2023年



