Bayes-DIC Net: Estimating Digital Image Correlation Uncertainty with Bayesian Neural Networks

2025-12-03

AI Summary

To address data scarcity and the lack of uncertainty quantification in displacement field prediction using Digital Image Correlation (DIC), this paper proposes Bayes-DIC Net. First, a high-fidelity, large-scale synthetic DIC dataset is generated using non-uniform B-splines. Second, a lightweight multi-scale convolutional network is designed, incorporating single-hop connectivity and optimized up/down-sampling for efficient feature extraction. Third, a Bayesian inference framework is integrated, where stochastic forward passes via activated dropout during inference enable predictive confidence estimation. This work presents the first end-to-end uncertainty-aware modeling for DIC, significantly enhancing model reliability and generalizability. Comprehensive evaluations on both synthetic and real-world experiments demonstrate consistently high-accuracy displacement predictions alongside well-calibrated uncertainty estimates.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models
• Computer Vision: Motion & Tracking

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

This paper introduces a novel method for generating high-quality Digital Image Correlation (DIC) dataset based on non-uniform B-spline surfaces. By randomly generating control point coordinates, we construct displacement fields that encompass a variety of realistic displacement scenarios, which are subsequently used to generate speckle pattern datasets. This approach enables the generation of a large-scale dataset that capture real-world displacement field situations, thereby enhancing the training and generalization capabilities of deep learning-based DIC algorithms. Additionally, we propose a novel network architecture, termed Bayes-DIC Net, which extracts information at multiple levels during the down-sampling phase and facilitates the aggregation of information across various levels through a single skip connection during the up-sampling phase. Bayes-DIC Net incorporates a series of lightweight convolutional blocks designed to expand the receptive field and capture rich contextual information while minimizing computational costs. Furthermore, by integrating appropriate dropout modules into Bayes-DIC Net and activating them during the network inference stage, Bayes-DIC Net is transformed into a Bayesian neural network. This transformation allows the network to provide not only predictive results but also confidence levels in these predictions when processing real unlabeled datasets. This feature significantly enhances the practicality and reliability of our network in real-world displacement field prediction tasks. Through these innovations, this paper offers new perspectives and methods for dataset generation and algorithm performance enhancement in the field of DIC.

Problem

• Generates realistic displacement datasets for DIC training
• Proposes a network for multi-level feature extraction in DIC
• Estimates prediction uncertainty using Bayesian neural networks in DIC

Innovation

• Generates DIC dataset using non-uniform B-spline surfaces
• Proposes Bayes-DIC Net with multi-level information extraction
• Transforms network into Bayesian for uncertainty estimation