Learn More
Falls are identified as a major health risk not only for the elderly but also for people with neurodegenerative diseases, such as epilepsy, and are considered as a major obstacle to independent living. Fast detection of falls would not only decrease the health risks by enabling quick medical response; but also make independent living a safe option for the(More)
We address the detection of vehicles in a video stream obtained from a moving airborne platform. Our approach is based on robust optical flow algorithm applied on stabilized frames. Stabilization of the frames compensates for gross affine background motion prior to running robust optical flow to compute dense residual flow. Based on the flow and the(More)
  • Hulya Yalcin
  • 2016
Human activity recognition using new generation depth sensors are particularly important for application that require human activity recognition. In this paper, a deep learning based algorithm is developed human activity recognition using RGB-D video sequences. Based on the assumption that every human activity is composed of many smaller actions, a temporal(More)
This paper describeV %89ø6-an automated visual inspection system – which selectively processes the incoming image data DQG FRPELQHV YLVLRQ ZLWK EHKDYLRU)RU WKLV %89ø6 LV HQGRZHG with biologically motivated visual capabilites: oculomation, attention and spatio-temporal reasoning. The advantages of such a system are being real-time and robust while running on(More)
Object recognition is a central problem in computer vision. When objects are defined by boundary curves, they can be represented either explicitly or implicitly. Implicit polynomial (IP) equations have long been known to offer certain advantages over more traditional parametric methods. However, the lack of general procedures for obtaining IP models of(More)
We consider detection of moving ground vehicles in airborne sequences recorded by a thermal sensor with automatic gain control, using an approach that integrates dense optic flow over time to maintain a model of background appearance and a foreground occlusion layer mask. However, the automatic gain control of the thermal sensor introduces rapid changes in(More)
Segmenting image sequences into meaningful layers is fundamental to many applications such as surveillance, tracking, and video summarization. Background subtraction techniques are popular for their simplicity and, while they provide a dense (pixelwise) estimate of foreground/background, they typically ignore image motion which can provide a rich source of(More)
A cognitive robot may face failures during the execution of its actions. These failures are mostly due to the gap between the physical world and the constructed symbolic plans, some internal problems that may occur in its embodiment or unexpected external events. In this paper, we propose a visual scene interpretation system for extracting spatial relations(More)
The aim of this paper is to help the communication of two people, one hearing impaired and one visually impaired by converting speech to fingerspelling and fingerspelling to speech. Fingerspelling is a subset of sign language, and uses finger signs to spell letters of the spoken or written language. We aim to convert finger spelled words to speech and vice(More)
Falls are identified as a major health risk not only for the elderly but also for people with cognitive diseases and are considered as a major obstacle to independent living. Fast detection of falls would not only decrease the health risks by enabling quick medical response; but also make independent living a safe option for the elderly. In this paper, we(More)