Neural Network Training System for Marker Encoding
- Авторы: Wang L.1, Grinchuk O.2, Tsurkov V.2,3
-
Учреждения:
- Nanjing University of Aeronautics and Astronautics
- Moscow Institute of Physics and Technology
- Federal Research Center Computer Science and Control, Russian Academy of Sciences
- Выпуск: Том 58, № 3 (2019)
- Страницы: 434-440
- Раздел: Artificial Intelligence
- URL: https://journals.rcsi.science/1064-2307/article/view/220387
- DOI: https://doi.org/10.1134/S1064230719030195
- ID: 220387
Цитировать
Аннотация
In this paper, we propose a training system for visual markers that provides the generation and subsequent recognition (under real-world conditions) of stylized images that contain information encoded by a sequence of bits. New types of neural network layers that make the recognition process tolerant to external noise are developed. The training process is based on the end-to-end principle, which enables automatic training for the intermediate stages of the model. The experimental results that demonstrate the performance of the system in encoding and decoding artificial markers are presented.
Об авторах
L. Wang
Nanjing University of Aeronautics and Astronautics
Email: tsurkov@ccas.ru
КНР, Nanjing
O. Grinchuk
Moscow Institute of Physics and Technology
Автор, ответственный за переписку.
Email: oleg.grinchuk@phystehc.edu
Россия, Dolgoprudny, 141701
V. Tsurkov
Moscow Institute of Physics and Technology; Federal Research Center Computer Science and Control, Russian Academy of Sciences
Автор, ответственный за переписку.
Email: tsurkov@ccas.ru
Россия, Dolgoprudny, 141701; Moscow, 119991
![](/img/style/loading.gif)