In this paper an efficient method of small object localization is proposed that integrates detection and tracking. The system is initialized using a strong detector and then it lo...
Sign language recognition constitutes a challenging field of research in computer vision. Common problems like overlap, ambiguities, and minimal pairs occur frequently and require...
Motion information is a powerful cue for visual perception. In the context of video indexing and retrieval, motion content serves as a useful source for compact video representati...
This paper presents a complete solution to estimating a scene’s 3D geometry and appearance from multiple 2D images by using a statistical inverse ray tracing method. Instead of ...
In this article we define a multimedia content analysis problem, which we call multimodal location estimation: Given a video/image/audio file, the task is to determine where it wa...