Object-Based Navigation Using Salient Landmarks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional navigation systems rely on map-based instructions, which can be difficult for users to follow when street signs are missing or occluded, especially in circuitous areas, and do not naturally align with human navigation methods that utilize salient landmarks.
Innovation Solution
A system that uses deep neural networks to identify salient objects within a user's field of view, providing navigational instructions based on these objects, such as landmarks or vehicles, to improve navigation accuracy and intuitiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If map-based navigation instructions are provided with reference to road names and absolute distances, then navigational information can be provided systematically, but users may have difficulty resolving road names and distances when street signs are missing or occluded
Solution Approach 1:
The patent introduces salient objects (landmarks, vehicles, pedestrians) as intermediary reference points between the navigation system and the user. Instead of directly referencing road names and distances, the system identifies and presents prominent objects in the user's field of view that can serve as natural landmarks for navigation instructions, making it easier for users to locate themselves and follow directions even when traditional street signs are missing or occluded
Solution Approach 2:
The patent replaces the traditional mechanical map-based reference system with an image-based object recognition system. By using cameras to capture the user's field of view and deep neural networks to identify salient objects, the system substitutes the abstract map coordinate system with concrete visual landmarks that users can naturally perceive and use for orientation
2Measurement precision
If conventional navigation systems use absolute distance measurements, then navigational instructions can be provided precisely, but users may struggle to comprehend and resolve these measurements in circuitous areas
Solution Approach 1:
The patent creates a visual copy of the user's actual field of view and superimposes identified salient objects onto it. This allows users to see the recognized landmarks directly in their camera view, creating a bridge between the abstract distance measurements and the concrete visual environment, making it easier to comprehend spatial relationships and distances in circuitous areas
3Ease of operation
If deep neural networks are used to identify salient objects, then navigational instructions become more intuitive and easier to follow, but computational complexity increases
Solution Approach 1:
The patent applies partial action by not attempting to identify all objects in the scene, but only those that are salient and relevant for navigation. The deep neural network focuses on detecting and identifying only the most prominent objects that would serve as effective landmarks, rather than performing exhaustive object detection on every element in the user's field of view, thus reducing computational complexity while maintaining ease of operation
Data Source
AI summary
In an aspect of the disclosure, a method, a computer-readable medium, and apparatus for operating a computational network are provided. The apparatus may obtain a set of navigational instructions describing a route to a destination. The apparatus may obtain first image data through a first camera that is oriented toward the route, the first image data depicting a first scene associated with the route. The apparatus may determine a first field of view associated with a user that is navigating the route to the destination based on a first sensor that is oriented toward the user. The apparatus may identify at least one salient object represented in the first scene based on the first field of view. The apparatus may output instructional information describing a first navigational instruction of the set of navigational instructions with reference to the at least one salient object.