Idea
Diffusion-based holographic reconstruction platform recovering amplitude and phase from intensity data for biomedical and imaging researchers
Research Paper
Core Innovation
This paper introduces a diffusion model trained only on amplitude data to reconstruct both amplitude and phase in holography without ground-truth phase training. It uses a predictor-corrector sampling framework with separate likelihood gradients for amplitude and phase, enabling complex field recovery. This approach generalizes well across diverse objects and imaging setups, including lensless systems.
Market Size (TAM)
$2–10B TAM for computational imaging and holography software; $1–3B SAM from biomedical imaging and industrial inspection sectors. Driven by demand for improved phase retrieval and cost-effective imaging solutions.
Potential Customers & Pain Points
- Biomedical Researchers Needing Accurate Phase Retrieval
- Computational Imaging Developers Seeking Generalizable Models
- Optical System Designers Requiring Robust Reconstruction
- Medical Imaging Labs Using Lensless or Coherent Imaging
- Industrial Inspection Teams Facing Nonlinear Inverse Problems
Business Model
SaaS platform offering API access to holographic reconstruction models; licensing to imaging hardware manufacturers; custom solutions for biomedical and industrial clients
Competitive Landscape
- PhaseFocus
- LightField Imaging
- Holoxica
Implementation Challenges
- Integration with existing imaging hardware
- Computational complexity of diffusion models
- Adoption resistance due to new methodology
Validation Strategy
- Conduct pilot studies with biomedical imaging labs
- Benchmark against existing phase retrieval methods
- Demonstrate robustness across diverse imaging modalities
Research Paper Overview
Generalizable Holographic Reconstruction via Amplitude-Only Diffusion Priors
Summary
Phase retrieval in inline holography is a fundamental yet ill-posed inverse problem due to the nonlinear coupling between amplitude and phase in coherent imaging. We present a novel off-the-shelf solution that leverages a diffusion model trained solely on object amplitude to recover both amplitude and phase from diffraction intensities. Using a predictor-corrector sampling framework with separate likelihood gradients for amplitude and phase, our method enables complex field reconstruction without requiring ground-truth phase data for training. We validate the proposed approach through extensive simulations and experiments, demonstrating robust generalization across diverse object shapes, imaging system configurations, and modalities, including lensless setups. Notably, a diffusion prior trained on simple amplitude data (e.g., polystyrene beads) successfully reconstructs complex biological tissue structures, highlighting the method's adaptability. This framework provides a cost-effective, generalizable solution for nonlinear inverse problems in computational imaging, and establishes a foundation for broader coherent imaging applications beyond holography.