Multi-Sensor Object Recognition via Probabilistic Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user terminal devices face challenges in accurately and quickly recognizing objects due to ambiguous natural language inputs and environmental noise, which can lead to distorted commands and incorrect information.
Innovation Solution
A user terminal device equipped with multiple sensors (sound, smell, camera, position, bio-signal, etc.) that acquire and analyze various types of data to determine object categories and estimate the most probable object based on probabilistic numbers, integrating information from multiple sources for precise recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If voice recognition technology is used to convert natural language into commands, then user input capability is improved, but recognition accuracy deteriorates due to ambiguous meaning and homonyms
Solution Approach 1:
The patent combines multiple sensing units (voice recognition, image recognition, smell recognition, position sensing) into an integrated object recognition system. By merging these different sensing modalities, the system overcomes the limitations of voice-only recognition and achieves more accurate object identification through cross-validation of multiple data sources.
Solution Approach 2:
The patent introduces an object recognition module as an intermediary between the user input and the execution module. This intermediary processes and validates information from multiple sensing units before generating commands, thereby improving recognition accuracy by filtering out ambiguous interpretations.
2Measurement precision
If multiple sensors are activated for object recognition, then recognition accuracy is improved, but device complexity increases
Solution Approach 1:
The patent designs a controller that performs multiple functions: it manages various sensing units (voice, image, smell, position), processes their outputs, and generates commands. This multi-functional controller reduces overall system complexity by consolidating control logic rather than requiring separate processing units for each sensor type.
Solution Approach 2:
The patent segments the object recognition system into distinct functional modules: voice recognition unit, image recognition unit, smell recognition unit, position sensing unit, and an object recognition module that integrates their outputs. This modular segmentation allows each component to be optimized independently while maintaining overall system manageability.
3Productivity
If multiple sensing units are used for object recognition, then recognition speed is improved, but energy consumption increases
Solution Approach 1:
The patent activates multiple sensing units partially or selectively based on the recognition task requirements. Rather than continuously operating all sensors at full capacity, the system activates only the necessary sensing units for each specific recognition scenario, thereby maintaining fast recognition speed while reducing overall energy consumption.
Data Source
AI summary
A user terminal device is provided. The user terminal device includes a plurality of different types of sensors, a storage unit configured to store object information, and a controller configured to acquire sensing results of the plurality of sensors, determine a plurality of object categories respectively corresponding to the sensing results, and compare objects belonging to the plurality of object categories to estimate an object commonly belonging to the object categories as a recognition result.