UDC 528.8
UDC 630
This work aims to determine the feasibility of using consumer artificial intelligence (neural networks), such as ChatGPT and Deepseek, for accelerated analysis of data in various formats. The data of interest are satellite images of terrain and text documents. The ability of such artificial intelligence systems to analyze text is unquestionable; not a single homework assignment has been written using them. The results of analyzing terrain images using these neural networks were, as expected, unsatisfactory.
satellite imagery analysis, neural networks, artificial intelligence, forest cover, automation
1. Ashish Vaswani, Noam Shazeer / Attention Is All You Need // 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA
2. Isaac Caswell, Bowen Liang / Recent Advances in Google Translate // Google Research June 8, 2020.
3. Pisarev A.E., Semenyuk O.I. / Opredelenie veroyatnosti provedeniya yadernyh ispytaniy s primenenime neyronnyh setey // Aktual'nye voprosy ekspluatacii sistem ohrannogo monitoringa i zaschischennyh telekommunikacionnyh sistem. Sbornik materialov Vserossiyskoy nauchno-prakticheskoy konferencii ad'yunktov, kursantov i slushateley. Voronezh, 2023. S. 397-398.
4. Google Earth. URL: https://earth.google.com/web/@51.31039722,42.14922149,137.12083673a,47426.68764185d,35y,357.77923466h,0.26896084t,0r/data=CgRCAggBOgMKATBCAggASg0I____________ARAA (data obrascheniya 10.09.2025).
5. Deepseek. URL: https://www.deepseek.com/en (data obrascheniya 10.09.2025).
