Quantifying the impact of non-stationarity in reinforcement learning-based traffic signal control.

In reinforcement learning (RL), dealing with non-stationarity is a challenging issue. However, some domains such as traffic optimization are inherently non-stationary. Causes for and effects of this are manifold. In particular, when dealing with traffic signal controls, addressing non-stationarity is key since traffic conditions change over time and as a function of traffic control decisions taken in other parts of a network. In this paper we analyze the effects that different sources of non-stationarity have in a network of traffic signals, in which each signal is modeled as a learning agent. More precisely, we study both the effects of changing the context in which an agent learns (e.g., a change in flow rates experienced by it), as well as the effects of reducing agent observability of the true environment state. Partial observability may cause distinct states (in which distinct actions are optimal) to be seen as the same by the traffic signal agents. This, in turn, may lead to sub-optimal performance. We show that the lack of suitable sensors to provide a representative observation of the real state seems to affect the performance more drastically than the changes to the underlying traffic patterns.

Detecting cells in intravital video microscopy using a deep convolutional neural network.

The analysis of leukocyte recruitment in intravital video microscopy (IVM) is essential to the understanding of inflammatory processes. However, because IVM images often present a large variety of visual characteristics, it is hard for an expert human or even conventional machine learning techniques to detect and count the massive amount of cells and extract statistical measures precisely. Convolutional neural networks are a promising approach to overcome this problem, but due to the difficulty of labeling cells, large data sets with ground truth are rare. The present work explores an adaptation of the RetinaNet model with a suite of augmentation techniques and transfer learning for detecting leukocytes in IVM data. The augmentation techniques include simulating the Airy pattern and motion artifacts present in microscopy imaging, followed by traditional photometric, geometric and smooth elastic transformations to reproduce color and shape changes in cells. In addition, we analyzed the use of different network backbones, feature pyramid levels, and image input scales. We have found that even with limited data, our strategy not only enables training without overfitting but also boosts generalization performance. Among several experiments, the model reached a value of 94.84 for the average precision (AP) metric as our best outcome when using data from different image modalities. We also compared our results with conventional image processing techniques and open-source tools. The results showed an outstanding precision of the method compared with other approaches, presenting low error rates for cell counting and centroid distances. Code is available at: https://github.com/brunoggregorio/retinanet-cell-detection.

Author Correction: Optical Absorption Exhibits Pseudo-Direct Band Gap of Wurtzite Gallium Phosphide.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Optical Absorption Exhibits Pseudo-Direct Band Gap of Wurtzite Gallium Phosphide.

Definitive evidence for the direct band gap predicted for Wurtzite Gallium Phosphide (WZ GaP) nanowires has remained elusive due to the lack of strong band-to-band luminescence in these materials. In order to circumvent this problem, we successfully obtained large volume WZ GaP structures grown by nanoparticle-crawling assisted Vapor-Liquid-Solid method. With these structures, we were able to observe bound exciton recombination at 2.14 eV with FHWM of approximately 1 meV. In addition, we have measured the optical absorption edges using photoluminescence excitation spectroscopy. Our results show a 10 K band gap at 2.19 eV and indicate a weak oscillator strength for the lowest energy band-to-band absorption edge, which is a characteristic feature of a pseudo-direct band gap semiconductor. Furthermore, the valence band splitting energies are estimated as 110 meV and 30 meV for the three highest bands. Electronic band structure calculations using the HSE06 hybrid density functional agree qualitatively with the valence band splitting energies.

Exploring Au Droplet Motion in Nanowire Growth: A Simple Route toward Asymmetric GaP Morphologies.

Here we show a new nanowire growth procedure, exploring the thermally activated motion of Au droplets on III-V surfaces. We show that by setting a single growth parameter we can activate the crawling motion of Au droplets in vacuum and locally modify surface composition in order to enhance vapor-solid (VS) growth along oxide-free areas on the trail of the metal particle. Asymmetric VS growth rates are comparable in magnitude to the vapor-liquid-solid growth, producing unconventional wurtzite GaP morphologies, which shows negligible defect density as well as optical signal in the green spectral region. Finally, we demonstrate that this effect can also be explored in different substrate compositions and orientations with the final shape finely tuned by group III flow and nanoparticle size. This distinct morphology for wurtzite GaP nanomaterials can be interesting for the design of nanophotonics devices.

Physical-chemical characteristics and potential use of a novel alginate/zein hydrogel as the sorption phase for polar organic compounds.

A novel alginate based hydrogel was successfully prepared and tested for the solid phase microextraction of medium-to-high polarity compounds, when supported in a polypropylene (PP) fiber. Pure alginate when added onto the surface of the PP fiber, resulted in a significant improvement in the extraction efficiency of the analytes (except for ß-estradiol). The alginate hydrogel was modified upon the incorporation of a small amount of zein, a corn protein. Interestingly, the alginate/zein-supported hydrogel was capable of successfully extracting compounds with low partition constant (Kow), such as 17-α-ethinyl estradiol, progesterone and estriol, since the initial water uptake decreased dramatically in this gel, therefore, leaving the alginate hydroxyl groups more available to interact with the polar compounds. In conclusion, this paper presents the preparation of a simple, low cost, reusable, and efficient sorption phase for the extraction of polar compounds with different polarities in aqueous samples, which is a current technological challenge in developing efficient wastewater treatment.

Automatic detection of motion blur in intravital video microscopy image sequences via directional statistics of log-Gabor energy maps.

Intravital microscopy is an important experimental tool for the study of cellular and molecular mechanisms of the leukocyte-endothelial interactions in the microcirculation of various tissues and in different inflammatory conditions of in vivo specimens. However, due to the limited control over the conditions of the image acquisition, motion blur and artifacts, resulting mainly from the heartbeat and respiratory movements of the in vivo specimen, will very often be present. This problem can significantly undermine the results of either visual or computerized analysis of the acquired video images. Since only a fraction of the total number of images are usually corrupted by severe motion blur, it is necessary to have a procedure to automatically identify such images in the video for either further restoration or removal. This paper proposes a new technique for the detection of motion blur in intravital video microscopy based on directional statistics of local energy maps computed using a bank of 2D log-Gabor filters. Quantitative assessment using both artificially corrupted images and real microscopy data were conducted to test the effectiveness of the proposed method. Results showed an area under the receiver operating characteristic curve (AUC) of 0.95 (AUC = 0.95; 95 % CI 0.93-0.97) when tested on 329 video images visually ranked by four observers.

Extraction optimization and antinociceptive activity of (1→3)-ß-d-glucan from Rhodotorulamucilaginosa.

ß-d-glucans are polymers of d-glucose monomers found in the cell walls of many bacteria, plants, fungi and yeasts. A variety of ß-d-glucans differing in structures have been isolated from various sources and their biological activity to be regulated by various structural factors, such as the primary structure, molecular weight, solubility, and conformation. This study investigated the effect of extraction time and temperature on the yield of ß-d-glucan produced by Rhodotorulamucilaginosa. A statistical Doehlert design was applied to determine the important effects and interactions of these independent variables on the yield of ß-d-glucan, the dependent variable. Significant models were obtained. The best yield was of 25% obtained after 128min of extraction in a temperature of 72°C. The polysaccharides were characterized as (1⟶3)-ß-d-glucan by methods spectroscopic (FT-IR, (1)HNMR and (13)CNMR). In addition, the antinociceptive effect was evaluated using different experimental tests (acetic acid-induced writhing test, formalin test and tail immersion test). The (1⟶3)-ß-d-glucan showed a potent peripheral antinociceptive effect, possibly by the inhibition of inflammatory mediators.

Structural characterization and emulsifying properties of polysaccharides of Acacia mearnsii de Wild gum.

Polysaccharides (GNF) from Acacia mearnsii de Wild gum exudates, collected from trees growing in the south of Brazil, were characterized ((13)C and HSQC NMR, GC-MS, colorimetric assays). A commercial gum arabic (GAC) was analyzed similarly and compared with GNF. There were differences, consistent with distinct behavior in tensiometry tests and as emulsion stabilizer. GNF had a higher protein content than GAC, with small differences in the monosaccharide composition, the greater one being the lower uronic acid content of GNF (4%), compared with GAC (17%). GNF had a much broader molecular mass distribution, M(w)/M(n), and a lower M(w). GNF was more efficient in lowering the surface tension of water and saline solutions and was more efficient in emulsifying castor oil droplets. Results were discussed taking into account structural and molecular differences between the studied gums. It was concluded that polysaccharides from A. mearnsii de Wild are candidates as substitutes of currently commercialized arabic gums (Acacia senegal and Acacia seyal) having, depending on their application, improved properties.