Computer Vision Asset Valuation Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the value of assets are inefficient and inaccurate, particularly when users lack expertise in asset valuation, and rely on user-input values that can lead to errors and resource wastage.
Innovation Solution
The use of computer vision and machine learning techniques to identify asset types and determine their values, with a novel machine learning approach to train models for accurate asset valuation, and the selective use of machine learning-based techniques based on asset type and accuracy thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning-based valuation is applied to all assets, then valuation accuracy may be improved, but computational resource consumption increases
Solution Approach 1:
The patent segments the asset valuation process into two distinct paths: a fast valuation path for assets suitable for automated machine learning valuation, and a manual valuation path for assets requiring human expertise. This segmentation allows the system to apply computational resources selectively rather than uniformly across all assets, thereby reducing overall resource consumption while maintaining accuracy for appropriate asset types.
Solution Approach 2:
The system applies different valuation qualities to different asset types based on their characteristics. For asset types where machine learning has been trained and proven effective, automated valuation is applied with high accuracy. For asset types outside the trained scope or requiring nuanced judgment, manual valuation by experts is invoked. This local quality approach ensures optimal resource utilization matched to specific asset requirements.
2Productivity
If automated machine learning valuation is used for all assets, then processing speed is improved, but valuation accuracy deteriorates for asset types outside training data
Solution Approach 1:
The system dynamically selects the valuation approach based on the specific asset being evaluated. The asset type determination model automatically routes each asset to the appropriate valuation path (automated or manual) based on its characteristics. This dynamic adaptation allows the system to maintain high processing speed for suitable assets while ensuring accuracy for complex or unfamiliar asset types by invoking manual valuation only when necessary.
Solution Approach 2:
The system incorporates feedback mechanisms where the asset type determination model learns from the characteristics of assets that require manual valuation. This feedback loop allows the system to progressively improve its routing decisions, increasing the proportion of assets that can be accurately valued through automated means while maintaining the safety net of manual valuation for edge cases.
3Measurement precision
If manual valuation methods are used, then accuracy may be maintained for complex assets, but processing time and resource wastage increase
Solution Approach 1:
The system performs preliminary asset type determination and classification before committing to a full valuation process. By quickly assessing whether an asset falls within the scope of trained machine learning models, the system can pre-route appropriate assets to automated valuation, avoiding the time-consuming manual process for those cases. This preliminary action filters out the majority of routine assets from manual review, significantly reducing overall processing time while preserving manual valuation for only those assets that truly require it.
Data Source
AI summary
A processing platform may receive a plurality of images. The processing platform may determine respective asset types of the plurality of assets based on a computer vision technique. The processing platform may determine respective estimated values of the plurality of assets based on the respective asset types. The processing platform may provide information identifying the respective estimated values of the plurality of assets to two or more recipients. The processing platform may receive allocation information. The processing platform may determine a selected allocation of the plurality of assets for the two or more recipients based on the allocation information and using a second model. The processing platform may perform one or more actions based on the selected allocation.