摘要
The ever-increasing complexity of robot applications induces the need for methods to approach problems with no (viable) analytical solution. Deep learning (DL) provides a set of tools to address this kind of problems. This survey presents a categorization of the major challenges in robotics that leverage DL technologies and introduces representative examples of successful solutions for the described problems. We also consider the question when and whether to use modular, monolithic models or end-to-end DL, in order to provide a guideline for the selection of the correct model structure and training strategy. By doing so, the current role and adaptability of different techniques at different hierarchical levels of a robot-application can be highlighted, thus providing a well-structured basis to assist future approaches.
学科与关键词
- Towards Deep Developmental LearningSigaud, Olivier*; Droniou, AlainIEEE Transactions on Cognitive and Developmental Systems, 2016, 8(2): 99-114.被引 24
- The path of foundational research in roboticsTrenton, JerdeNature Machine Intelligence, 2020, 2(3): 153-154.
- Deep learning: Applications, architectures, models, tools, and frameworks: A comprehensive surveyGheisari, Mehdi; Ebrahimzadeh, Fereshteh; Rahimi, Mohamadtaghi; Moazzamigodarzi, Mahdieh; Liu, Yang*; Pramanik, Pijush Kanti Dutta; Heravi, Mohammad Ali; Mehbodniya, Abolfazl; Ghaderzadeh, Mustafa; Feylizadeh, Mohammad Reza*; Kosari, SaeedCAAI Transactions on Intelligence Technology, 2023, 8(3): 581-606.被引 81
