An optimisation of Gaussian mixture models for integer processing units

This paper investigates sub-integer implementations of the adaptive Gaussian mixture model (GMM) for background/foreground segmentation to allow the deployment of the method on low cost/low power processors that lack Floating Point Unit (FPU). We propose two novel integer computer arithmetic techniques to update Gaussian parameters. Specifically, the mean value and the variance of each Gaussian are updated by a redefined and generalised "round'' operation that emulates the original updating rules for a large set of learning rates. Weights are represented by counters that are updated following stochastic rules to allow a wider range of learning rates and the weight trend is approximated by a line or a staircase. We demonstrate that the memory footprint and computational cost of GMM are significantly reduced, without significantly affecting the performance of background/foreground segmentation.

Sher 25: pulsating but apparently alone

The blue supergiant Sher 25 is surrounded by an asymmetric, hourglass-shaped circumstellar nebula, which shows similarities to the triple-ring structure seen around SN 1987A. From optical spectroscopy over six consecutive nights, we detect periodic radial velocity variations in the stellar spectrum of Sher 25 with a peak-to-peak amplitude of ~ 12 km s^-1 on a time-scale of about 6 d, confirming the tentative detection of similar variations by Hendry et al. From consideration of the amplitude and time-scale of the signal, coupled with observed line profile variations, we propose that the physical origin of these variations is related to pulsations in the stellar atmosphere, rejecting the previous hypothesis of a massive, short-period binary companion. The radial velocities of two other blue supergiants with similar bipolar nebulae, SBW1 and HD 168625, were also monitored over the course of six nights, but these did not display any significant radial velocity variations. 

Ecology of aspergillosis: insights into the pathogenic potency of Aspergillus fumigatus and some other Aspergillus species

Fungi of the genus Aspergillus are widespread in the environment. Some Aspergillus species, most commonly Aspergillus fumigatus, may lead to a variety of allergic reactions and life-threatening systemic infections in humans. Invasive aspergillosis occurs primarily in patients with severe immunodeficiency, and has dramatically increased in recent years. There are several factors at play that contribute to aspergillosis, including both fungus and host-related factors such as strain virulence and host pulmonary structure/immune status, respectively. The environmental tenacity of Aspergilllus, its dominance in diverse microbial communities/habitats, and its ability to navigate the ecophysiological and biophysical challenges of host infection are attributable, in large part, to a robust stress-tolerance biology and exceptional capacity to generate cell-available energy. Aspects of its stress metabolism, ecology, interactions with diverse animal hosts, clinical presentations and treatment regimens have been well-studied over the past years. Here, we synthesize these findings in relation to the way in which some Aspergillus species have become successful opportunistic pathogens of human- and other animal hosts. We focus on the biophysical capabilities of Aspergillus pathogens, key aspects of their ecophysiology and the flexibility to undergo a sexual cycle or form cryptic species. Additionally, recent advances in diagnosis of the disease are discussed as well as implications in relation to questions that have yet to be resolved.