In this paper we propose DFL -- a formal, graphical workflow language for dataflows, i.e., workflows where large amounts of complex data are manipulated, and the structure of the ...
Jan Hidders, Natalia Kwasnikowska, Jacek Sroka, Je...
This paper presents a novel evolutionary approach for function optimization Incremental Evolution Strategy (IES). Two strategies are proposed. One is to evolve the input variables...
In this paper a method for extraction of mid-level semantics from sign language videos is proposed, by employing high level domain knowledge. The semantics concern labeling of the ...
In this paper we show that the initial philosophy used in designing and developing UNIX in early times has been forgotten due to “fast practices”. We question the leitmotif th...
Results of neural network learning are always subject to some variability, due to the sensitivity to initial conditions, to convergence to local minima, and, sometimes more dramat...