SLAM is a challenging problem in robotics, involving building a map of an unknown environment while simultaneously localizing a robot within it. Both conventional and neural methods have been used for SLAM, with the choice depending on factors such as the application, available computational resources, and required accuracy. Conventional methods have been successful for many years, but neural methods are gaining popularity for their ability to handle uncertainty and learn from data. The choice between these methods depends on the problem at hand, and hybrid approaches that consider the strengths and weaknesses of each have shown the best recent results.
Махмуд Ж. A REVIEW AND COMPARISON BETWEEN CONVENTIONAL AND NEURAL SLAM ALGORITHMS // Сборник тезисов докладов конгресса молодых ученых. Электронное издание. – СПб: Университет ИТМО, [2023]. URL: https://kmu.itmo.ru/digests/article/10314