Towards Robust and Generalizable Gerchberg Saxton based Physics Inspired Neural Networks for Computer Generated Holography: A Sensitivity Analysis Framework

2025-04-30

AI Summary

In computer-generated holography (CGH), phase retrieval—a critical inverse problem—has long suffered from poor generalization and hardware adaptability due to neural networks’ high sensitivity to the physical forward model (FM) and its hyperparameters (FMHs). Method: We propose the first systematic sensitivity analysis framework for Gerchberg–Saxton physics-informed neural networks (GS-PINNs), employing Saltelli-extended Sobol’ indices to quantify FMH influence, designing a composite metric integrating consistency, generalization, and robustness to perturbations, and deriving interpretable guidelines for FM selection and network architecture design. Results: Experiments identify SLM pixel resolution as the dominant sensitivity factor; the free-space propagation model substantially outperforms the Fourier hologram model. Our framework establishes a unified benchmark, significantly enhancing GS-PINN robustness and cross-hardware generalizability across multiple wavelengths, propagation distances, and pixel sizes.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Probabilistic Circuits and Graphical Models
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Computer-generated holography (CGH) enables applications in holographic augmented reality (AR), 3D displays, systems neuroscience, and optical trapping. The fundamental challenge in CGH is solving the inverse problem of phase retrieval from intensity measurements. Physics-inspired neural networks (PINNs), especially Gerchberg-Saxton-based PINNs (GS-PINNs), have advanced phase retrieval capabilities. However, their performance strongly depends on forward models (FMs) and their hyperparameters (FMHs), limiting generalization, complicating benchmarking, and hindering hardware optimization. We present a systematic sensitivity analysis framework based on Saltelli's extension of Sobol's method to quantify FMH impacts on GS-PINN performance. Our analysis demonstrates that SLM pixel-resolution is the primary factor affecting neural network sensitivity, followed by pixel-pitch, propagation distance, and wavelength. Free space propagation forward models demonstrate superior neural network performance compared to Fourier holography, providing enhanced parameterization and generalization. We introduce a composite evaluation metric combining performance consistency, generalization capability, and hyperparameter perturbation resilience, establishing a unified benchmarking standard across CGH configurations. Our research connects physics-inspired deep learning theory with practical CGH implementations through concrete guidelines for forward model selection, neural network architecture, and performance evaluation. Our contributions advance the development of robust, interpretable, and generalizable neural networks for diverse holographic applications, supporting evidence-based decisions in CGH research and implementation.

Problem

• Addressing phase retrieval challenges in computer-generated holography (CGH)
• Analyzing forward model hyperparameter impacts on GS-PINN performance
• Establishing robust evaluation metrics for CGH neural networks

Innovation

• Sensitivity analysis framework for GS-PINNs
• Composite metric for benchmarking CGH configurations
• Guidelines for robust CGH neural networks