Digital evolution is a computer-based instantiation of Darwinian evolution in which short self-replicating computer programs compete, mutate, and evolve. It is an excellent platform for addressing topics in long-term evolution and paleobiology, such as mass extinction and recovery, with experimental evolutionary approaches. We evolved model communities with ecological interdependence among community members, which were subjected to two principal types of mass extinction: a pulse extinction that killed randomly, and a selective press extinction involving an alteration of the abiotic environment to which the communities had to adapt. These treatments were applied at two different strengths, along with unperturbed control experiments. We examined how stability in the digital communities was affected from the perspectives of division of labor, relative shift in rank abundance, and genealogical connectedness of the community's component ecotypes. Mass extinction that was due to a Strong Press treatment was most effective in producing reshaped communities that differed from the pre-treatment ones in all of the measured perspectives; weaker versions of the treatments did not generally produce significant departures from a Control treatment; and results for the Strong Pulse treatment generally fell between those extremes. The Strong Pulse treatment differed from others in that it produced a slight but detectable shift towards more generalized communities. Compared to Press treatments, Pulse treatments also showed a greater contribution from re-evolved ecological doppelgangers rather than new ecotypes. However, relatively few Control communities showed stability in any of these metrics over the whole course of the experiment, and most did not represent stable states (by some measure of stability) that were disrupted by the extinction treatments. Our results have interesting, broad qualitative parallels with findings from the paleontological record, and show the potential of digital evolution studies to illuminate many aspects of mass extinction and recovery by addressing them in a truly experimental manner.

No abstract

Content Based Image Retrieval is very hottest research area in computer vision and image processing. To perceive arbitrary natural scene from complex environment is a challenging issue in visual imaging and processing research area. Neural Network is a grid of "neuron like" nodes, in this paper we follow towards Neural Network (NN), is committed to contributing a new technical concept for the scene understanding and recognition by consolidating new intellectual visual features into the scene expression, which can be very crucial and provide cognitive intelligence to cloud robot. Inspired by Artificial Neural Network intelligence due to its dynamic nature, we make use of the attributes of the Gabor filter and Laplacian of Gaussian filter which is to be akin to robot visual perception, and apply the wavelet transform to inspect a new approach in complex environment natural scene perception and understanding for virtual phenomena. Through the study of Neural Network, the perception ability of the natural scene image from complex environment for cloud robot is enhanced with the integration of cognitive visual features and the scene expression.

No abstract

No abstract