Dr. Yozevitch's Research Fields
My Main interest is applying machine learning and deep learning in order to solve problems in the Humanities realm. Among others, incorporating computer vision and DL methods to classify motoric-cognitive abilities among children and youth. Another research aims to detect neural synchrony states by utilizing data obtained from a depth camera. My PhD dealt with GNSS accuracy improvement algorithms, mainly in dense urban environments. On this work I won the prestigious Wolf Scholarship, given each year to the most dozen promising students in Israel. One can read more in the KC&G lab (Kinematic and Computational Geometry) or visit my Google Scholar Page.
GNSS Accuracy improvement
For urban canyon GNSS accuracy improvement, the KCG lab is a serious source of knowledge with serval patents in the field. Commercial GNSS devices tend to perform poorly in urban canyon environments. The dense and tall buildings block the signals from many of the satellites. In the KCG lab, we developed a particle filter algorithm for the Shadow-Matching framework to face this problem.
Given a 3D city map and given the satellites’ signal properties, the algorithm calculates in real-time invalid regions inside the Region Of Interest (ROI). This approach reduces the ROI to a fraction of its original size. We present a general framework for Shadow Matching positioning algorithm based on a modified particle filter. Using simulation experiments we have shown that the suggested method can improve the accuracy of existing GNSS devices in urban regions. Moreover, the proposed algorithm can be efficiently extended to 3D positioning in a high sampling rate, inherently applicable for UAVs and Drones.
cognitive-motoric assessments using A.I.
This study deals with the creation of a learning system based on hybrid NN (image + metadata) for an objective, accurate and rapid analysis of motor-cognitive abilities. The idea is to provide an assessment of these abilities in the test taker (children and adults) by analyzing the forms they copy on a tablet using a smartpen.
Among the innovations of the proposed system is the use of metrics that have not been available to researchers so far, in both quality and quantity. Among other things, the strength and grip of the pen throughout the test at a resolution unavailable to a human examiner.
The research is conducted at Bar-Ilan University with Professor Rachel Sheiff.
As part of my research in the Kinematic and computational Geometry lab, I was also involved in the "Satlla project" is the most pretentious research project ever conducted In the KCG lab. The realm of nanosatellite is an ever-growing field in the industry. Alas, most nanosatellites rely on commercial of-the-self modules. In the KCG lab, we think differently.
First, we designed and built the entire electronics of the satellite. Thus, the system is much more flexible to any modifications that need to be done.
Second, the Sattla 1.0 works with several new communication modules: the commercial APRS (UHV/VHV), Iridium, STX satellite modem by globalsat, Lora 2.4, and Lora 433Mhz.
Third, the satellite is equipped with 40 high-amp LEDs that can transmit serial communication in rate up to 1 Mhz. This FSO (Free Space Optics) communication can open a new field of swarm satellite optic communication.
Fourth, one of the modules is a Spectro gamma device designed to record electromagnetic storm in the LEO height (200-600 km). This research will enable far more accurate predictions of geo-events such as earthquakes, tsunamis, etc.