Cyclist Orientation Estimation Using LiDAR Data.

It is crucial for an autonomous vehicle to predict cyclist behavior before decision-making. When a cyclist is on real traffic roads, his or her body orientation indicates the current moving directions, and his or her head orientation indicates his or her intention for checking the road situation before making next movement. Therefore, estimating the orientation of cyclist's body and head is an important factor of cyclist behavior prediction for autonomous driving. This research proposes to estimate cyclist orientation including both body and head orientation using deep neural network with the data from Light Detection and Ranging (LiDAR) sensor. In this research, two different methods are proposed for cyclist orientation estimation. The first method uses 2D images to represent the reflectivity, ambient and range information collected by LiDAR sensor. At the same time, the second method uses 3D point cloud data to represent the information collected from LiDAR sensor. The two proposed methods adopt a model ResNet50, which is a 50-layer convolutional neural network, for orientation classification. Hence, the performances of two methods are compared to achieve the most effective usage of LiDAR sensor data in cyclist orientation estimation. This research developed a cyclist dataset, which includes multiple cyclists with different body and head orientations. The experimental results showed that a model that uses 3D point cloud data has better performance for cyclist orientation estimation compared to the model that uses 2D images. Moreover, in the 3D point cloud data-based method, using reflectivity information has a more accurate estimation result than using ambient information.

DOMINANT SPECIES OF THE GENUS PROTOPERIDINIUM BERGH (PERIDINIALES: PROTOPERIDINIACEAE) IN THE BLACK SEA.

To date, 59 species (61 including infraspecific taxa) of the heterotrophic genus Protoperidinium have been found in the Black Sea. With rare exceptions, the records of most species were not confirmed by either line drawings or microphotographs, and lacked indications of important diagnostic traits. To partially correct these deficiencies, we conducted morphological examinations of the dominant Protoperidinium species in the Black Sea. The study provides original descriptions, line drawings, and microphotographs of 22 species of Protoperidinium with their global distribution maps. Protoperidinium quadrioblongum is reported from the Black Sea for the first time.

Pedestrian Detection Using Multispectral Images and a Deep Neural Network.

Pedestrian fatalities and injuries most likely occur in vehicle-pedestrian crashes. Meanwhile, engineers have tried to reduce the problems by developing a pedestrian detection function in Advanced Driver-Assistance Systems (ADAS) and autonomous vehicles. However, the system is still not perfect. A remaining problem in pedestrian detection is the performance reduction at nighttime, although pedestrian detection should work well regardless of lighting conditions. This study presents an evaluation of pedestrian detection performance in different lighting conditions, then proposes to adopt multispectral image and deep neural network to improve the detection accuracy. In the evaluation, different image sources including RGB, thermal, and multispectral format are compared for the performance of the pedestrian detection. In addition, the optimizations of the architecture of the deep neural network are performed to achieve high accuracy and short processing time in the pedestrian detection task. The result implies that using multispectral images is the best solution for pedestrian detection at different lighting conditions. The proposed deep neural network accomplishes a 6.9% improvement in pedestrian detection accuracy compared to the baseline method. Moreover, the optimization for processing time indicates that it is possible to reduce 22.76% processing time by only sacrificing 2% detection accuracy.

Motion planning strategies in human control of non-rigid objects with internal degrees of freedom.

The paper deals with modeling of human-like reaching movements in dynamic environments. A simple but not trivial example of reaching in a dynamic environment is the rest-to-rest manipulation of a multi-mass flexible object with the elimination of residual vibrations. Two approaches to the prediction of reaching movements are formulated in position and force actuation settings. In the first approach, either the position of the hand or the hand force is specified by the lowest order polynomial satisfying the boundary conditions of the reaching task. The second approach is based on the minimization of either the hand jerk or the hand force-change, with taking into account the dynamics of the flexible object. To verify the resulting four mathematical models, an experiment on the manipulation of a ten-masses flexible object of low stiffness is conducted. The experimental results show that the second approach gives a significantly better prediction of human movements, with the minimum hand force-change model having a slight but consistent edge over the minimum hand jerk one.

Electric field sensing with liquid-crystal-filled slot waveguide microring resonators.

We consider the operation principles of the sensor of an external electric field on the bases of microring resonators that consist of bent-strip waveguides with vertical or horizontal slots filled with nematic liquid crystal. The mode-field distribution and dispersion parameters of the bent-slot waveguides are calculated by using the algorithm based on the method of lines. The influence of the waveguide and microresonator structure (slot width, position and orientation, microresonator radius, etc.) on the sensor sensitivity is analyzed.

Design of High-Specificity Nanocarriers by Exploiting Non-Equilibrium Effects in Cancer Cell Targeting.

Although targeting of cancer cells using drug-delivering nanocarriers holds promise for improving therapeutic agent specificity, the strategy of maximizing ligand affinity for receptors overexpressed on cancer cells is suboptimal. To determine design principles that maximize nanocarrier specificity for cancer cells, we studied a generalized kinetics-based theoretical model of nanocarriers with one or more ligands that specifically bind these overexpressed receptors. We show that kinetics inherent to the system play an important role in determining specificity and can in fact be exploited to attain orders of magnitude improvement in specificity. In contrast to the current trend of therapeutic design, we show that these specificity increases can generally be achieved by a combination of low rates of endocytosis and nanocarriers with multiple low-affinity ligands. These results are broadly robust across endocytosis mechanisms and drug-delivery protocols, suggesting the need for a paradigm shift in receptor-targeted drug-delivery design.

Vicious walks with long-range interactions.

The asymptotic behavior of the survival or reunion probability of vicious walks with short-range interactions is generally well studied. In many realistic processes, however, walks interact with a long-ranged potential that decays in d dimensions with distance r as r(-d-σ). We employ methods of renormalized field theory to study the effect of such long-range interactions. We calculate the exponents describing the decay of the survival probability for all values of parameters σ and d to first order in the double expansion in ε=2-d and δ=2-d-σ. We show that there are several regions in the σ-d plane corresponding to different scalings for survival and reunion probabilities. Furthermore, we calculate the leading logarithmic corrections.

Vicious Lévy flights.

We study the statistics of encounters of Lévy flights by introducing the concept of vicious Lévy flights--distinct groups of walkers performing independent Lévy flights with the process terminating upon the first encounter between walkers of different groups. We show that the probability that the process survives up to time t decays as t-α at late times. We compute α up to the second order in ε expansion, where ε=σ-d, σ is the Lévy exponent, and d is the spatial dimension. For d=σ, we find the exponent of the logarithmic decay exactly. Theoretical values of the exponents are confirmed by numerical simulations. Our results indicate that walkers with smaller values of σ survive longer and are therefore more effective at avoiding each other.

Pressure-induced hydrogen-dominant metallic state in aluminum hydride.

Two structural transitions in covalent aluminum hydride AlH3 were characterized at high pressure. A metallic phase stable above 100 GPa is found to have a remarkably simple cubic structure with shortest first-neighbor H-H distances ever measured except in H2 molecule. Although the high-pressure phase is predicted to be superconductive, this was not observed experimentally down to 4 K over the pressure range 120-164 GPa. The results indicate that the superconducting behavior may be more complex than anticipated.

Evidence for a magnetic collapse in the epsilon phase of solid oxygen.

Solid oxygen is the only elementary molecular magnet. Under the very high pressure of 96 GPa oxygen transforms into a metal and a superconductor. Theory predicts a nonmagnetic state occurring before the transition into the superconducting xi phase. Nevertheless, until now there was no direct evidence of a magnetic collapse in high-pressure oxygen. For the first time direct information is provided on magnetic properties of the epsilon phase, which is sandwiched between the antiferromagnetic delta phase and the superconducting xi phase. We used magnetic neutron diffraction. The data show that the long-range magnetic order disappears at the delta-epsilon transition. The magnetic collapse occurs at P approximately equal to 8 GPa, far below the pressure of the insulator-metal (superconductor) transition. The collapse is preceded by a decrease in temperature of transition towards the long-range magnetically ordered state (T(LRO)) in the delta phase, at P = 7.6 GPa.

Neutron and X-ray diffraction study of the broken symmetry phase transition in solid deuterium.

The solid hydrogen compounds D2, HD and H2 remain quantum molecular solids up to pressures in the 100 GPa range. A remarkable macroscopic consequence is the existence of a pressure-induced broken symmetry phase transition, in which the molecules go from a spherical rotational state to an anisotropic rotational state. Theoretical understanding of the broken symmetry phase structure remains controversial, despite numerous studies. Some open questions concern the existence of long- or short-range orientational order; whether a strong isotopic shift on the transition pressure should be assigned to the nuclear zero-point motion or to quantum localization; and whether the structures are cubic, hexagonal or orthorhombic. Here we present experimental data on the structure of the broken symmetry phase in solid D2, obtained by a combination of neutron and X-ray diffraction up to 60 GPa. Our data are incompatible with orthorhombic structures predicted by recent theoretical works. We find that the broken symmetry phase structure is incommensurate with local orientational order, being similar to that found in metastable cubic para-D2.

Pressure-induced magnetization in FeO: evidence from elasticity and Mössbauer spectroscopy.

The complete elastic tensor of Fe0.94O (wüstite) has been determined to 10 GPa using acoustic interferometry at GHz frequencies inside a diamond-anvil cell. The soft mode (C44) elastic constant of FeO is reduced by 20% over the experimental pressure range. An unusual discontinuity in the pressure derivatives of C11 and C12 at 4.7+/-0.2 GPa corresponds to the pressure at which the onset of a magnetic ordering transition is observed by high-pressure Mössbauer spectroscopy and neutron powder diffraction. Our new results combined with literature structural high P-T data suggest that there is a magnetic, although still cubic, phase of FeO between approximately 5 and approximately 17 GPa.