Voice Guidance Landmark Prioritization for Foreign Visitors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems for foreign visitors to Japan, particularly those using voice guidance, struggle to provide effective landmark recognition due to reliance on translated characters and unfamiliar signage, leading to confusion and potential loss, especially when encountering new or unfamiliar shops and landmarks.
Innovation Solution
A guidance text generation device and system that generates and outputs voice guidance using easily recognizable landmarks such as global brands, universal signs, and objects peculiar to Japan, prioritizing geographical information based on ease of recognition, and updates guidance in real-time based on the user's location and orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional navigation systems use translated characters and unfamiliar signage for foreign visitors, then the system can provide basic navigation functionality, but the landmarks become difficult to recognize and users may get lost
Solution Approach 1:
The patent applies local quality by selecting landmarks with different recognition characteristics for different user groups. For foreign visitors, the system prioritizes landmarks with English signage or internationally recognizable symbols, while for local visitors, Japanese landmarks are emphasized. This creates localized guidance content tailored to each user's linguistic and cultural background, improving landmark recognizability without sacrificing navigation accuracy.
Solution Approach 2:
The system changes the parameter of landmark selection criteria based on user attributes. When detecting a foreign visitor, the system adjusts the selection parameters to prioritize landmarks with English signage, alphanumeric notation, or international recognition. This dynamic parameter adjustment ensures that the recommended landmarks match the user's ability to recognize and identify them, resolving the contradiction between ease of operation and navigation accuracy.
2Ease of operation
If navigation systems use Collaborative Filtering to estimate recognizable landmarks based on user attributes, then landmark recognition improves, but a large amount of data is required for learning and new landmarks cannot be recognized until relearned
Solution Approach 1:
The patent applies preliminary action by pre-classifying landmarks into categories based on their recognition characteristics (English signage, alphanumeric notation, international symbols, etc.) before users arrive. This pre-processing of landmark data allows the system to immediately recommend appropriate landmarks when a foreign visitor is detected, without requiring time-consuming data collection and machine learning. New landmarks can be quickly integrated by categorizing them according to their visual and linguistic characteristics.
Solution Approach 2:
The system segments the landmark database into distinct categories based on recognition characteristics such as signage language, symbol type, and international recognizability. This segmentation allows the navigation system to quickly filter and recommend landmarks from specific categories based on user attributes, eliminating the need for comprehensive machine learning of all landmarks. The segmented structure enables rapid adaptation to new landmarks through simple categorization rather than full relearning.
3Object-affected harmful factors
If voice navigation is used instead of screen-based navigation for walking, then safety improves by avoiding terminal operation while walking, but the complexity of providing multilingual and context-aware guidance increases
Solution Approach 1:
The patent applies universality by designing a voice navigation system that handles multiple functions through a unified process. The same voice synthesis engine that provides navigation instructions also delivers landmark information, turning directions, and contextual assistance. The system universally applies attribute-based landmark selection across all voice outputs, whether for route guidance or landmark identification, simplifying the overall system architecture while providing multilingual and context-aware guidance.
Solution Approach 2:
The system uses an intermediary classification layer that translates user attributes and landmark characteristics into appropriate guidance content. This intermediary process categorizes landmarks based on recognition features and matches them with user profiles, then feeds the results to the voice synthesis system. This intermediary step decouples the complexity of multilingual, context-aware landmark selection from the voice output mechanism, making the overall system more manageable while maintaining high functionality.
Data Source
AI summary
A guidance text generation device that generates a guidance text for a user includes a route generation unit that generates a route including nodes from a point of departure to a destination, the nodes being represented by the point of departure, corners or/and ends at which a traveling direction changes and the destination and geographical information including types for classifying things located on a route connecting nodes, the types being classified into at least one of global brands, universal signs/facilities, objects peculiar to Japan, shops/facilities with alphanumeric notation, and shops, facilities and objects that do not fall under any of such categories and a guidance text generation unit that generates a guidance text for the generated route based on the generated route, the geographical information on the generated route, presentation priority of the geographical information associated with the type of the geographical information.