Time-series are ordered sequences of discrete-time data. Due to their temporal dimension, anomaly detection techniques used in time-series have to take into consideration time correlations and other time-related particularities. Generally, in order to evaluate the quality of an anomaly detection technique, the confusion matrix and its derived metrics such as precision and recall are used. These metrics, however, do not take this temporal dimension into consideration. In this paper, we propose three metrics that can be used to evaluate the quality of a classification, while accounting for the temporal dimension found in time-series data.

No abstract

No abstract

Depth mapping can be carried out by ultrasound measuring devices using the time of flight method. Ultrasound measurements are favorable in such environments, where the light or radio frequency measurements can not be applied due to the noise level, calculation complexity, reaction time, size and price of the device, accuracy or electromagnetic compatibility. It is also usual to apply and fusion ultrasound sensors with other types of sensors to increase the precision and reliability. In the case of visually impaired people, an echolocation based aid for determining the distance and the direction of obstacles in the surroundings can improve the life quality by giving the possibility to move alone and individually in unlearnt or rapidly changing environments. In the following considerations, a model system is presented which can provide rather reliable position and distance to multiple objects. The mathematical model based on the time of flight method with a correction: it uses the measured analog sensor signals, which represent the probability of the presence of an obstacle. The device consists of multiple receivers, but only one source. The sensor fusion algorithm for this setup and the results of indoor experiments are presented. The mathematical model allows the usage, processing, and fusion of the signals of up to an infinite number of sensors. In addition, the positions of the sensors can be arbitrary, and the mathematical model does not restrict them to be placed in regular formations.

No abstract