Uncertain Object Detection/Segmentation

The uncertain scene object refers to the object which brings uncertainty to the scene understanding process for intelligence systems. The presence of uncertain objects in the scene severely affects intelligent decisions in many applications such as robotic navigation and drone tracking. In this project, we are developing techniques for uncertain object detection/segmentation. Based on environmental data perception, we are committed to mining the features and patterns of scene data from different perspectives such as computational statistics, behavioral cognition, and semantics, combining multi-dimensional and multi-modal technical means, so as to realize effective analysis, cognition, and expression of the uncertain scene object. Specifically, the uncertain objects we focus on include special, camouflaged, and salient objects in the scene.

{{project.title}} {{tag[0]}} [{{link.name}}] 

{{img.label}}

Input-Output: {{project.io}}

Abstract {{project.abstract}}

{{index+1}}