A Survey on Autonomous Mobile Navigation Using Multimodal Agents and Natural Language Commands
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Smartphones are mostly controlled by touch.However, current voice assistants provide limited navigation, especially for complex or changing interfaces.This limitation drives research into multimodal systems that combine speech, vision, and language understanding.Past work in this area falls into three categories: rule-based, OCR-based, and large language model (LLM)-driven methods.This survey reviews and compares these approaches, focusing on their methods, strengths, and weaknesses.It identifies major research gaps, such as adapting to changing user interfaces, improving multimodal perception, and the absence of unified benchmarks.The survey also points out ongoing challenges like latency, privacy, and real-world scalability, particularly in hands-free and accessibility situations.By summarizing recent advancements and comparing key systems, this work offers a helpful reference for researchers and practitioners looking to improve autonomous mobile navigation.
Fields of Study & Subjects
Speech and dialogue systems
Usability and User Interface Design
Multimodal Machine Learning Applications
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 174–190
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter