Search

This paper presents an approach to correcting misclassified pixels in depth maps representing parts of the human body. A misclassified pixel is a pixel of a depth map which, incorrectly, has the ‘background’ value and does not accurately reflect the distance from the sensor to the body being scanned. A completely automatic, deep learning based solution for depth map correction is proposed. As an input, the solution requires a color image and a corresponding erroneous depth map. The input color image is segmented using deep neural network for human body segmentation. The extracted segments are further used as guidance to find and amend the misclassified pixels on the depth map using a simple average based filter. Unlike other depth map refinement solutions, this paper designs a method for the improvement of the input depth map in terms of completeness instead of precision. The proposed method does not exclude the application of other refinement methods. Instead, it can be used as the first step in a depth map enhancement pipeline to determine approximate depths for erroneous pixels, while other refinement methods can be applied in a second step to improve the accuracy of the recovered depths.

This article analyzes the sensitivity of deep learning methods for ocular fundus segmentation. We use an empirical methodology based on non‐adversarial perturbed datasets. The research is motivated by mass screening and self‐administered tests in which AI methods are needed and may be given images with focus issues. These substandard pictures are simulated using blurring algorithms of varying designs and kernel sizes which are subjected to a test of inter‐network and inter‐dataset sensitivity using Gaussian noise as a control. We test for simulated defocusing, motion blur, and Gaussian noise. The DRIVE, CHASE, and STARE datasets were used to generate 441 synthetic datasets for testing. Analysis of the resultant n = 37,888 sample has identified anomalies illuminating failure modes of state‐of‐the‐art ocular segmentation models. Data analysis led us to conclude that accuracy is a poor measure of sensitivity to input change and that even moderate levels of blur have dramatic effects on said sensitivity. We further concluded that architectures show larger variations in sensitivity to blur which do not match either their reported or measured performance on unblurred test datasets. Our analysis has attributed at least a part of the problem to overfitting non‐essential input dataset features and resolution sensitivity. [ABSTRACT FROM AUTHOR]

Power quality disturbances (PQD) have a negative impact on power quality-sensitive equipment, often resulting in great financial losses. To prevent these losses, besides detecting a PQD on time, it is important to classify it, so that appropriate recovery procedures are employed. The majority of research employs machine learning model PQD classifiers on manually extracted features from simulated or real-world signals. This paper presents an end-to-end approach that circumvents the manual feature extraction and uses signals generated from mathematical voltage sag type formulas. We developed a configurable voltage sag generator that was used to form training and validation datasets. Based on the synthetic three-phase voltage signals, we trained several end-to-end LSTM classifiers that classify voltage sags according to ABC classification. The best-performing model achieved an accuracy of over 90% in the real-world dataset. [ABSTRACT FROM AUTHOR]