A Large-Depth-Range Layer-Based Hologram Dataset for Machine Learning-Based 3D Computer-Generated Holography

2025-12-24

AI Summary

To address the scarcity of high-quality, large-scale hologram datasets in machine learning–based computer-generated holography (ML-CGH), this work introduces KOREATECH-CGH—the first deep learning–oriented layered hologram dataset—comprising 6,000 RGB-D image pairs and their corresponding complex-valued holograms, spanning resolutions from 256×256 to 2048×2048 and covering the theoretical maximum depth range. We propose an amplitude-projection post-processing technique that adaptively replaces wavefield amplitudes across depth layers while preserving phase information, significantly improving reconstruction fidelity for large-depth scenes. Holograms are synthesized via the angular spectrum method, integrated with layered depth modeling and rigorously evaluated using PSNR and SSIM metrics. The dataset and method are validated on generative and super-resolution ML models. Our approach achieves 27.01 dB PSNR and 0.87 SSIM—outperforming state-of-the-art silhouette masking by 2.03 dB and 0.04, respectively—establishing a new benchmark dataset and an effective reconstruction paradigm for ML-CGH.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Computational Photography, Image & Video Synthesis
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Machine learning-based computer-generated holography (ML-CGH) has advanced rapidly in recent years, yet progress is constrained by the limited availability of high-quality, large-scale hologram datasets. To address this, we present KOREATECH-CGH, a publicly available dataset comprising 6,000 pairs of RGB-D images and complex holograms across resolutions ranging from 256*256 to 2048*2048, with depth ranges extending to the theoretical limits of the angular spectrum method for wide 3D scene coverage. To improve hologram quality at large depth ranges, we introduce amplitude projection, a post-processing technique that replaces amplitude components of hologram wavefields at each depth layer while preserving phase. This approach enhances reconstruction fidelity, achieving 27.01 dB PSNR and 0.87 SSIM, surpassing a recent optimized silhouette-masking layer-based method by 2.03 dB and 0.04 SSIM, respectively. We further validate the utility of KOREATECH-CGH through experiments on hologram generation and super-resolution using state-of-the-art ML models, confirming its applicability for training and evaluating next-generation ML-CGH systems.

Problem

• Addresses lack of large-scale hologram datasets for machine learning
• Improves hologram quality at large depth ranges via amplitude projection
• Validates dataset for training and evaluating ML-based holography systems

Innovation

• Introduces amplitude projection for hologram quality enhancement
• Creates large-scale hologram dataset with wide depth ranges
• Validates dataset with machine learning models for holography