Deep Learning-Driven Gaussian Modeling and Improved Motion Detection Algorithm of the Three-Frame Difference Method

To enhance the effect of motion detection, a Gaussian modeling algorithm is proposed to fix holes and breaks caused by the conventional frame difference method. The proposed algorithm uses an improved three-frame difference method. A three-frame image sequence with one frame interval is selected for pairwise difference calculation. The logical “OR” operation is used to achieve fast motion detection and to reduce voids and fractures. The Gaussian algorithm establishes an adaptive learning model to make the size and contour of the motion detection more accurate. The motion extracted by the improved three-frame difference method and Gaussian model is logically summed to obtain the final motion foreground picture. Moreover, a moving target detection method, based on the U-Net deep learning network, is proposed to reduce the dependency of deep learning on the number of training datasets. It helps the algorithm to train models on small datasets. Next, it calculates the ratio of the number of positive and negative samples in the dataset and uses the reciprocal of the ratio as the sample weight to deal with the imbalance of positive and negative samples. Finally, a threshold is set to predict the results for obtaining the moving object detection accuracy. Experimental results show that the algorithm can suppress the generation and rupture of holes and reduce the noise. Also, it can quickly and accurately detect movement to meet the design requirements.

CHANGES IN GAUSSIAN MODELING PARAMETERS OF PPG PULSE DURING HEALTHY PREGNANCY

We used Gaussian modeling to depict the changes in finger photoplethysmographic (PPG) pulse during pregnancy in healthy women. We enrolled 70 healthy pregnant women and recorded their PPG pulses in 11–13 gestational weeks, 20–22 gestational weeks, and 37–39 gestational weeks. Three independent positive Gaussian functions were utilized to decompose the pulses, and each Gaussian function extracted three key parameters: the peak amplitude ([Formula: see text]), the peak position ([Formula: see text]), and the half-width ([Formula: see text]). The method of ANOVA and post-hoc multiple comparisons of mathematical statistics were utilized to study the differences of these parameters between the three trimesters. We found that in the first trimester [Formula: see text] increased significantly ([Formula: see text]: [Formula: see text] versus [Formula: see text], [Formula: see text]). [Formula: see text] and [Formula: see text] increased in the first trimester ([Formula: see text]: [Formula: see text] versus [Formula: see text], [Formula: see text]; [Formula: see text]: [Formula: see text] versus [Formula: see text], [Formula: see text]), then decreased significantly ([Formula: see text]: [Formula: see text] versus [Formula: see text], [Formula: see text]: [Formula: see text] versus [Formula: see text], [Formula: see text]). [Formula: see text] is associated with cardiac output, and [Formula: see text] and [Formula: see text] are associated with peripheral vascular resistance. The results of this study were consistent with the conclusion that healthy pregnant women exhibited high flow state of the cardiovascular system and their peripheral vascular resistance decreased first and then gradually recovered during pregnancy. This study indicated that PPG pulse could also reflect the changes in the maternal cardiovascular system during pregnancy.

Worst performance rule, or not-best performance rule?

The worst performance rule (WPR) is a robust empirical finding reflecting that people’s worst task performance shows stronger relations to cognitive ability compared to their average or best performance. However, recent meta-analytic work has proposed this be renamed the “not-best-performance” rule because mean and worst performance seem to predict cognitive ability to similar degrees (with both predicting ability better than best performance). We re-analyzed data from a previously published latent-variable study to test for worst vs. not-best performance across a variety of reaction time tasks in relation to two cognitive ability constructs: working memory capacity (WMC) and task-unrelated thought (TUT) rate. Using two methods of assessing worst performance—ranked-binning and ex-Gaussian-modeling approaches—we found evidence for both worst and not-best performance rules. WMC followed the not-best performance rule (correlating equivalently with mean and worst RTs) but TUT propensity followed the worst performance rule (correlating more strongly with worst RTs). Additionally, we created a mini-multiverse following different outlier exclusion rules to test the robustness of our findings; our findings remained stable across the different multiverse iterations. We provisionally conclude that the worst performance rule may only arise in relation to cognitive abilities closely linked to (failures of) sustained attention.