Machine Learning Skin Color Estimation Under Uncontrolled Lighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately estimate skin color in images captured under uncontrolled lighting conditions, making it difficult to recommend beauty products like foundation online.
Innovation Solution
Training machine learning models to analyze images under varying lighting conditions, using normalization techniques and multiple images to improve accuracy, and applying these models to estimate skin color for product recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image processing methods are used to estimate skin color, then the process is simple and fast, but the accuracy is poor under uncontrolled lighting conditions
Solution Approach 1:
The system performs preliminary actions by capturing multiple images under different lighting conditions before making the final skin color estimation. This allows the machine learning model to learn from varied lighting scenarios and compensate for uncontrolled lighting conditions, thereby improving measurement precision without requiring complex hardware modifications
Solution Approach 2:
A machine learning model serves as an intermediary between the raw image data and the final skin color estimation. The model processes images captured under uncontrolled lighting conditions and predicts accurate skin color values, effectively mediating the transformation from inaccurate direct measurement to accurate inferred measurement
2Measurement precision
If multiple images under different lighting conditions are captured to improve accuracy, then skin color estimation accuracy improves, but the time required increases
Solution Approach 1:
Multiple images are captured in advance under different lighting conditions and stored for later processing. This preliminary action allows the system to have a rich dataset ready, and when skin color estimation is needed, the pre-captured images can be processed quickly by the machine learning model, reducing the actual processing time
Solution Approach 2:
The system continuously captures images under varying lighting conditions to build and update the training dataset. This continuous data collection ensures that the machine learning model is always working with fresh, diverse data, improving estimation accuracy while the automated processing maintains efficiency
3Adaptability or versatility
If machine learning models are trained with diverse training data to improve robustness, then the model generalizes better to uncontrolled conditions, but the training complexity and data requirements increase
Solution Approach 1:
The machine learning model is designed to be universal by training it on diverse images containing various skin tones, lighting conditions, and facial features. This multi-functional training approach enables the single model to handle different scenarios effectively, improving adaptability without requiring separate specialized models for each condition
Solution Approach 2:
The system uses automatically captured images from the environment to generate training data, eliminating the need for manual annotation and curation. The machine learning model trains itself on this self-generated diverse dataset, reducing the complexity of the training process while still achieving robustness to lighting variations
Data Source
AI summary
In some embodiments of the present disclosure, one or more machine learning models are trained to accurately estimate skin color in one or more images regardless of the lighting conditions. In some embodiments, the models can then be used to estimate a skin color in a new image, and that estimated skin color can be used for a variety of purposes. For example, the skin color may be used to generate a recommendation for a foundation shade that accurately matches the skin color, or a recommendation for another cosmetic product that is complimentary with the estimated skin color. Thus, the need for an in-person test of the product is eliminated.