Lipid tracking at kilohertz sampling rates on live cell membranes through Interferometric Scattering microscopy

The lateral dynamics of lipids on the cellular membranes are one of the most challenging topics to study in membrane biophysics, needing simultaneously high spatial and temporal resolution. In this study, we have employed Interferometric scattering Microscopy (ISCAT) to explore the dynamics of a biotinylated lipid analogue labelled with streptavidin-coated gold nanoparticles (20 and 40nm in diameter) at 2kHz sampling rate. We developed a statistics-driven analysis pipeline to analyse both ensemble average and single trajectory Mean Squared Displacements from each dataset, and to discern the most likely diffusion mode. We found that the use of larger tags slows down the target motion without affecting the diffusion mode. Moreover, we determined from our statistical analysis that the prevalent diffusion mode of the tracked gold-labelled lipids is compartmentalized diffusion. This model describes the motion of particles diffusing on a corralled surface, with a certain probability of changing compartment. This is compatible with the picket-fence model of membrane structure, already observed by similar studies. Through our analysis, we could determine significant physical parameters, such as average compartment size, dynamic localization uncertainty, and the intra- and inter-compartmental diffusion rates. We then simulated diffusion in an environment compatible with the experimentally-derived parameters and model. The closeness of the results from the analysis of experimental and simulated trajectories validates our analysis and the proposed description of the cell membrane. Finally, we introduce the confinement strength metric to compare diffusivity measurements across techniques and experimental conditions, which we used to successfully compare the present results with other related studies.

Optimal Trajectory Planning for Wheeled Mobile Robots under Localization Uncertainty and Energy Efficiency Constraints

With the rapid development of robotics, wheeled mobile robots are widely used in smart factories to perform navigation tasks. In this paper, an optimal trajectory planning method based on an improved dolphin swarm algorithm is proposed to balance localization uncertainty and energy efficiency, such that a minimum total cost trajectory is obtained for wheeled mobile robots. Since environmental information has different effects on the robot localization process at different positions, a novel localizability measure method based on the likelihood function is presented to explicitly quantify the localization ability of the robot over a prior map. To generate the robot trajectory, we incorporate localizability and energy efficiency criteria into the parameterized trajectory as the cost function. In terms of trajectory optimization issues, an improved dolphin swarm algorithm is then proposed to generate better localization performance and more energy efficiency trajectories. It utilizes the proposed adaptive step strategy and learning strategy to minimize the cost function during the robot motions. Simulations are carried out in various autonomous navigation scenarios to validate the efficiency of the proposed trajectory planning method. Experiments are performed on the prototype “Forbot” four-wheel independently driven-steered mobile robot; the results demonstrate that the proposed method effectively improves energy efficiency while reducing localization errors along the generated trajectory.

Face Pose Alignment with Event Cameras

Event camera (EC) emerges as a bio-inspired sensor which can be an alternative or complementary vision modality with the benefits of energy efficiency, high dynamic range, and high temporal resolution coupled with activity dependent sparse sensing. In this study we investigate with ECs the problem of face pose alignment, which is an essential pre-processing stage for facial processing pipelines. EC-based alignment can unlock all these benefits in facial applications, especially where motion and dynamics carry the most relevant information due to the temporal change event sensing. We specifically aim at efficient processing by developing a coarse alignment method to handle large pose variations in facial applications. For this purpose, we have prepared by multiple human annotations a dataset of extreme head rotations with varying motion intensity. We propose a motion detection based alignment approach in order to generate activity dependent pose-events that prevents unnecessary computations in the absence of pose change. The alignment is realized by cascaded regression of extremely randomized trees. Since EC sensors perform temporal differentiation, we characterize the performance of the alignment in terms of different levels of head movement speeds and face localization uncertainty ranges as well as face resolution and predictor complexity. Our method obtained 2.7% alignment failure on average, whereas annotator disagreement was 1%. The promising coarse alignment performance on EC sensor data together with a comprehensive analysis demonstrate the potential of ECs in facial applications.