Ryerson University

About: Ryerson University is a education organization based out in Toronto, Ontario, Canada. It is known for research contribution in the topics: Computer science & Population. The organization has 7671 authors who have published 20164 publications receiving 394976 citations. The organization is also known as: Ryerson Polytechnical Institute & Ryerson Institute of Technology.

Surface-active solid lipid nanoparticles as Pickering stabilizers for oil-in-water emulsions.

Abstract: Oil-in-water (O/W) emulsions solely stabilized by surface-active solid lipid nanoparticles (SLNs) were developed. The SLNs were generated by quench-cooling hot O/W nanoemulsions consisting of 7.5% glyceryl stearyl citrate (GSC) dispersed in water. Their initial volume-weighted mean particle diameter (∼152 nm) and zeta potential (ca. −49 mV) remained unchanged for 24 weeks. O/W emulsions (oil phase volume fraction: 0.2) containing 7.5% (w/w) GSC SLNs in the aqueous phase were kinetically-stable for 12 weeks and did not visually phase-separate over 24 weeks. The O/W emulsions generated with solid-state GSC SLNs had a volume-weighted mean oil droplet diameter of ∼459 nm and a zeta potential of ca. −43 mV. Emulsion microstructure evaluated with TEM revealed dispersed oil droplets sparsely covered with adsorbed Pickering-type SLNs as well aggregated SLNs present in the continuous phase. Gradual emulsion destabilization resulted from GSC SLN dissolution during the experimental timeframe. Overall, surface-active SLNs developed via nanoemulsions effectively kinetically stabilized O/W emulsions.

Selective Harmonic Elimination Model Predictive Control for Multilevel Power Converters

Abstract: In this study, a model predictive control (MPC) strategy that combines finite-control-set MPC with selective harmonic elimination (SHE) modulation pattern in its formulation is proposed to govern multilevel power converters. Based on a desired operating point for the system state (converter current reference), an associated predefined SHE voltage pattern is obtained as a required steady-state control input reference. Then, the cost function is formulated with the inclusion of both system state and control input references. According with the proposed reference and cost function formulation, the predictive controller prefers to track the converter output current reference in transients, while preserving the SHE voltage pattern in steady state. Hence, as evidenced by experimental results, a fast dynamic response is obtained throughout transients while a predefined voltage and current spectrum with low switching frequency is achieved in steady state.

Interactions of Vitamin D3 with Bovine β-Lactoglobulin A and β-Casein

Abstract: It is of nutritional significance to fortify processed dairy products (e.g., cheese, yogurt, and ice cream) with vitamin D3; however, the inherent complexity of these foods may influence the stability and bioavailability of this nutrient. In the present study, the interactions of vitamin D3 with β-lactoglobulin A and β-casein were investigated under various environmental conditions (i.e., pH and ionic strength) using fluorescence and circular dichroism spectroscopic techniques. The results indicated that vitamin D3 was bound to both β-lactoglobulin A and β-casein depending on the solution conditions. The apparent dissociation constants ranged from 0.02 to 0.29 μM for β-lactoglobulin A, whereas the β-casein apparent dissociation constants ranged from 0.06 to 0.26 μM. The apparent mole ratios were also comparable, i.e., 0.51−2.04 and 1.16−2.05 mol of vitamin D3 were bound per mole of β-lactoglobulin A and β-casein, respectively. It was concluded that these interactions may strongly influence vitamin D3 stab...

6-DoF Object Pose from Semantic Keypoints

Abstract: This paper presents a novel approach to estimating the continuous six degree of freedom (6-DoF) pose (3D translation and rotation) of an object from a single RGB image. The approach combines semantic keypoints predicted by a convolutional network (convnet) with a deformable shape model. Unlike prior work, we are agnostic to whether the object is textured or textureless, as the convnet learns the optimal representation from the available training image data. Furthermore, the approach can be applied to instance- and class-based pose recovery. Empirically, we show that the proposed approach can accurately recover the 6-DoF object pose for both instance- and class-based scenarios with a cluttered background. For class-based object pose estimation, state-of-the-art accuracy is shown on the large-scale PASCAL3D+ dataset.