RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

PNIPU: Points for analysis of customer needs using neural networks

Product
The name of the base system (platform): Artificial intelligence (AI, Artificial intelligence, AI)
Developers: PNIPU Perm National Research Polytechnic University
Date of the premiere of the system: 2021/10/15
Technology: Video analytics systems

Main article: Video analytics (terms, applications, technologies)

2021: Development of glasses, programs for "tracking" views and video analysis systems

The development of Perm Polytechnic will help to find out the needs of customers using neural networks. This was announced by the university on October 15, 2021.

The convenient location of the goods and their presentable appearance play an important role in attracting customers. To make the choice the most comfortable, companies use the technology of "tracking the direction of view." Programmers from Perm Polytechnic have created glasses based on neural network algorithms that will allow you to find out which products and interfaces most attract visitors. The product will help to create a complete picture of the client's behavior and will be cheaper than analogues by about 10 times, according to developers.

Product developers and marketers need to understand how customers interact with goods. For this, aitracking technologies are used - "tracking the direction of views."

File:Aquote1.png
Representatives of the focus group put on special glasses that allow you to find out in which direction and in what sequence customers look, what their look stops longer. This allows you to adjust the design of the interfaces or effectively position the goods in the store. But existing technologies are expensive and can only process data from one point. Therefore, the analysis process takes a long time, "said Alexander Knyazev, project developer, programmer, graduate student of the Faculty of Electrical Engineering of Perm Polytechnic.
File:Aquote2.png

Unlike analogues, domestic development will allow you to study information from several points at once. This was achieved through special algorithms based on neural networks that optimize the process of image processing, and the use of multi-core computing modules.

The attention analysis system includes several glasses with built-in cameras. Two of them are located under the eyes and monitor the look, and the third camera is between the eyes and shows the image that a person sees. Wi-Fi glasses transmit information to the server where data is processed and visualized. The analysis is performed by a system of two parts: glasses and software for them, as well as server and software for it. You can use the user's computer as the server.

The glasses software will be able to receive video from cameras and process it. Due to this, the direction of view will be displayed in the "picture" that the user sees. The program will be able to send data to the server.

Young researchers from Perm Polytechnic have already developed a prototype of glasses, a program for "tracking" views and a video analysis system.

According to developers, the product will be interesting for implementation in the b2b market, but the technology can be adapted to solve other problems. For example, with glasses, users can control a computer without a mouse.