We propose a principled framework to model persistent motion in dynamic scenes. In contrast to previous efforts on object tracking and optical flow estimation that focus on local...
In this paper, we deal with the problem of scheduling streaming applications on unreliable heterogeneous platforms. We use the realistic one-port model with full computation/commu...
Although the main fault-tolerance techniques are known for a long time, there exists no consistent approach for implementing dependable applications in the sense that the fault-to...
We report on the development of a visual programming environment for building applications for execution on a range of parallel computing platforms. This work exploits the dataflo...
Abstract. In order to take full advantage of high-end computing platforms, scientific applications often require modifications to source codes, and to their build systems that ge...