Numerical-modeling-based data acquiring capability
Multimodal-sensor-based data acquiring capability
AI model compatibility
Architecture of the CNN model for image-based prediction of properties for porous materials
Machine Learning Model for Saturated Soil Water Content Prediction
Machine Learning Model for Soil Dry Density Prediction
Data-driven by experiments and simulations
Fast, efficient, and accurate property estimation
Well-suited for multiscale, multiphase material research
Compatible with various ML architectures
Sensitivity and impact factor analysis supportive
Photo-based ML Property Prediction
CNN Model + RGB–XY decision tree
Bulk and regional thermal property prediction
Appliable for general porous/granular/multiphase materials