ML Imagery Suitability Scoring via Preliminary Action
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital imagery processing technologies lack efficient methods to identify and categorize imagery suitable for specific uses, such as particular applications or audiences, using machine learning models.
Innovation Solution
A computer-implemented method and system that utilize an application programming interface (API) to process imagery using machine learning (ML) models, determining scores for each frame indicating its suitability for different uses, and communicating these scores to requesting applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are used to process and analyze imagery data, then the precision of identifying suitable imagery is improved, but the computing resources and processing time required increase
Solution Approach 1:
The system performs preliminary actions by pre-processing imagery data and pre-computing features before the actual suitability assessment. ML models are trained in advance on large datasets, and feature extraction is performed beforehand, so that when actual imagery needs to be evaluated, the computationally intensive work has already been completed, reducing real-time computing resource requirements while maintaining high precision.
Solution Approach 2:
The imagery processing task is segmented into multiple independent stages: feature extraction, preprocessing, ML model inference, and suitability scoring. Each stage can be processed independently and potentially in parallel, allowing computing resources to be distributed and utilized more efficiently across different processing steps rather than requiring all resources simultaneously for a single monolithic operation.
2Reliability
If machine learning models process multiple frames of imagery, then the comprehensiveness of suitability assessment is improved, but the processing time increases
Solution Approach 1:
The system maintains continuous useful action by implementing a streamlined processing pipeline that continuously evaluates multiple frames without interruption. Once the ML model is trained and the processing framework is established, the system can efficiently process multiple frames in sequence or parallel, maintaining high comprehensiveness of assessment while minimizing idle time and overall processing duration through optimized data flow and computational efficiency.
Solution Approach 2:
The system applies partial action by evaluating only the most relevant features and aspects of each frame that are necessary for suitability assessment, rather than analyzing every possible attribute in exhaustive detail. The ML model is trained to focus on key discriminative features, allowing comprehensive assessment of suitability while processing only the essential information needed, thus reducing overall processing time.
3Ease of operation
If an API is provided for processing imagery using ML models, then the ease of operation for requesting applications is improved, but the device complexity increases
Solution Approach 1:
The API serves as an intermediary layer between requesting applications and the complex ML processing system. It provides a simplified, standardized interface that abstracts away the underlying complexity of the ML models, data preprocessing requirements, and computation logistics. Applications can make simple API calls to request imagery processing without needing to understand or implement the complex processing pipeline, thus improving ease of operation while the system complexity is contained and managed within the processing system itself.
Data Source
AI summary
The present disclosure is directed to processing imagery using one or more machine learning (ML) models. In particular, data describing imagery comprising a plurality of different and distinct frames can be received; and based at least in part on one or more ML models and the data describing the imagery, and for each frame of the plurality of different and distinct frames, one or more scores can be determined for the frame. Each score of the score(s) can indicate a determined measure of suitability of the frame with respect to one or more of various different and distinct uses for which the ML model(s) are configured to determine suitability of imagery.