Crowdsourced Location Platform Using Iterative Expectation-Maximization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current crowdsourcing platforms are inadequate for accurately locating targets of interest in images, as they either fail to provide precise location information or are prone to inaccuracies due to ambiguity and variability in user inputs, especially in search and locate problems that require precise geolocation of multiple targets.
Innovation Solution
A crowdsourced search and locate platform that utilizes a closed-loop analysis model, where user inputs are processed through a server-based system to receive and analyze image data, providing navigation and tagging functionality, and an algorithm that generates quality scores, difficulty scores, and estimated locations of targets, using iterative expectation-minimization processes to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If crowdsourced inputs are used for target location, then cost and scalability are improved, but measurement precision and reliability deteriorate due to ambiguity and variability in user inputs
Solution Approach 1:
The system implements an iterative expectation-maximization algorithm that processes crowdsourced location inputs through multiple cycles. In each iteration, the algorithm refines target location estimates by comparing predicted observations with actual user inputs, progressively improving accuracy while maintaining scalability. This feedback loop transforms noisy crowd inputs into reliable location data.
Solution Approach 2:
The patent introduces a server-based processing system that acts as an intermediary between crowdsourced user inputs and final target location results. This intermediary layer applies the expectation-maximization algorithm to filter, aggregate, and refine raw user inputs, converting variable crowd data into precise location estimates without requiring direct human analysis of each input.
2Ease of operation
If simple crowdsourcing methods are used, then ease of operation is improved, but measurement precision deteriorates due to inability to handle ambiguity in user inputs
Solution Approach 1:
The expectation-maximization algorithm provides iterative feedback processing that gradually resolves ambiguities in simple user inputs. The system starts with initial location estimates from crowd inputs and refines them through multiple calculation cycles, allowing the platform to maintain ease of operation while achieving high measurement precision through algorithmic refinement.
Solution Approach 2:
The system changes the state of location data from raw, ambiguous user inputs to refined, precise estimates through the expectation-maximization process. By transforming parameters (location coordinates, confidence levels) through iterative calculation, the system elevates simple inputs into accurate target location data without complicating the user interface.
3Productivity
If multiple user inputs are aggregated, then productivity is improved, but device complexity increases due to need for iterative processing algorithms
Solution Approach 1:
The server-based expectation-maximization algorithm acts as an intermediary processing layer that handles the complexity of aggregating and analyzing multiple user inputs. This centralized intermediary absorbs the computational complexity, allowing individual user devices to remain simple while the system as a whole achieves high productivity through sophisticated batch processing of crowd inputs.
Data Source
AI summary
A crowdsourced search and locate platform, comprising an application server that receives input from a plurality of user devices and navigates to a particular location, sends images of the location to a user device, and receives tagging data provided by the device user, and a client interface application that displays a plurality of interactive elements to a user, receive input from the user, and provide the results of the input to the application server, and methods for operating and administering a crowdsourced search and locate platform.