Information Processing Apparatus Selective Data Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In existing systems, it is difficult to effectively extract and utilize large amounts of classified data accumulated in databases, especially for machine learning applications in robots and automobiles, where past data classification is insufficient for effective use.
Innovation Solution
An information processing apparatus and method that sets labels based on element information in scene information, determining whether to store the information, thereby restricting data volume and enhancing usability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If large amounts of classified data are collected and accumulated in a database, then data availability is improved, but data extraction difficulty increases and usability decreases
Solution Approach 1:
The patent segments data into different categories by assigning labels to scene information. The labeling unit divides the continuous stream of scene data into discrete, manageable categories based on element information, making it easier to extract and query specific types of data from the large accumulated dataset.
Solution Approach 2:
The patent introduces labels as an intermediary between the raw scene information and the query requirements. These labels act as mediators that organize and index the data, enabling efficient retrieval of specific information without requiring direct search through the entire large database.
2Reliability
If all scene information is stored without selection, then data completeness is improved, but data management complexity and storage requirements increase
Solution Approach 1:
The patent applies local quality by selectively storing only the portions of scene information that are relevant and useful, rather than uniformly storing all data. The determination unit evaluates each piece of scene information and stores only those that meet specific criteria, reducing storage requirements and management complexity while maintaining data completeness for useful information.
Solution Approach 2:
The patent changes the parameter of data selection by using labels as a filtering mechanism. The determination unit uses label information to decide whether to store scene data, transforming the storage process from unconditional to conditional based on data characteristics and usefulness.
3Loss of information
If past data classification is used for machine learning, then historical information is preserved, but effectiveness for current applications decreases
Solution Approach 1:
The patent implements feedback by using the determination unit to continuously evaluate scene information and adjust what is stored based on current needs. This feedback mechanism ensures that historical data is preserved through labeling while the system adapts to current machine learning requirements by selectively storing data that will be most useful for future applications.
Solution Approach 2:
The patent introduces dynamics by making the data storage process adaptive rather than static. The determination unit dynamically decides what to store based on the characteristics of scene information and current machine learning needs, allowing the system to evolve from simple past data classification to effective current applications.
Data Source
AI summary
An information processing apparatus includes: a processor in communication with a memory configured to store instructions that, when executed by the processor, cause the processor to set a label based on element information included in scene information for a scene, and determine, based on the set label, whether or not to store the scene information set with the label.