Beyond the Thin-Layer Limit: Differentiable Volumetric Training for Visible-Range Diffractive Neural Networks

• URL: http://arxiv.org/abs/2606.07896v1
• Date: Fri, 05 Jun 2026 23:14:06 GMT
• Title: Beyond the Thin-Layer Limit: Differentiable Volumetric Training for Visible-Range Diffractive Neural Networks
• Abstract summary: Translating D2NNs to the visible range, where nearly all vision pipelines operate, was long blamed on the difficulty of fabricating nanoscale neurons.<n>We identify the true obstacle as the thin-layer approximation underlying nearly all D2NN training, which treats each diffractive layer as an infinitely thin mask.<n>We introduce a differentiable beam-propagation layer that models each element as a finite-thickness volume and propagates light through it during training, keeping the fabrication-compatible height map end-to-end trainable.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Diffractive deep neural networks (D2NNs) promise miniaturized, power-efficient, light-speed optical front-ends for machine vision, yet the most mature demonstrations remain in the terahertz regime, built from readily fabricated millimeter-scale neurons. Translating D2NNs to the visible range, where nearly all vision pipelines operate, was long blamed on the difficulty of fabricating nanoscale neurons; but even after recent advances removed that barrier, visible-range D2NNs matching their terahertz counterparts remain out of reach. We identify the true obstacle as the thin-layer approximation underlying nearly all D2NN training, which treats each diffractive layer as an infinitely thin mask. It fails not because of the short wavelength, as is commonly assumed, but because the low-refractive-index materials (n approximately 1.3-1.5) used at visible wavelengths require relief structures thick enough that intra-layer diffraction and phase accumulation become significant. To overcome this, we introduce a differentiable beam-propagation ($\partial$BPM) layer that models each element as a finite-thickness volume and propagates light through it during training, keeping the fabrication-compatible height map end-to-end trainable without full-wave simulation in the loop. Across MNIST, Fashion-MNIST, and CIFAR-100 classification and imaging, $\partial$BPM training substantially reduces the design-to-device mismatch, and full-wave FDTD validation raises classification accuracy from 50% to 90% without re-optimization. The $\partial$BPM layer thus offers a scalable, physics-aware bridge between efficient optical neural-network optimization and fabrication-consistent diffractive design.

Related papers

• Current Injection Spiking Neural Network for Infrared and Visible Image Fusion [55.20584003351512]
  Infrared and visible image fusion (IVIF) integrates the complementary information of two modalities into a single image with richer scene content.<n>We propose CIS-Fuse, a spiking network that performs cross-modal fusion directly at the membrane-potential level.<n>Experiments on four IVIF benchmarks and on downstream detection and segmentation show that CIS-Fuse achieves fusion quality on par with state-of-the-art ANN-based methods.
  arXiv  Detail & Related papers  (2026-07-22T08:10:02Z)
• DexWorldModel: Causal Latent World Modeling towards Automated Learning of Embodied Tasks [54.32016216994156]
  We introduce the Causal Latent World Model (CLWM), which employs DINOv3 features as generative targets to disentangle interaction semantics from visual noise.<n>To overcome memory scaling, CLWM features a Dual-State Test-Time Training (TTT) Memory that guarantees a strict $mathcalO(1)$ footprint for long-horizon tasks.<n>We present EmbodiChain, an online framework that establishes the Efficiency Law by injecting an infinite flow of physics-grounded trajectories during training.
  arXiv  Detail & Related papers  (2026-04-13T03:19:36Z)
• Brain-Inspired Multimodal Spiking Neural Network for Image-Text Retrieval [79.04396446873932]
  Spiking neural networks (SNNs) have recently shown strong potential in unimodal visual and textual tasks.<n>Building a directly trained, low-energy, and high-performance SNN for multimodal applications such as image-text retrieval (ITR) remains highly challenging.<n>We present a brain-inspired Cross-Modal Spike Fusion network (CMSF) and apply it to ITR for the first time.
  arXiv  Detail & Related papers  (2026-03-25T08:41:07Z)
• NanoSD: Edge Efficient Foundation Model for Real Time Image Restoration [5.158202521463481]
  NanoSD is a general-purpose diffusion foundation model family suitable for real-time visual generation and restoration on edge devices.<n>We show how architectural balance, feature routing, and latent-space preservation jointly shape true on-device latency.<n>When used as a drop-in backbone, NanoSD enables state-of-the-art performance across image super-resolution, image deblurring, face restoration, and monocular depth estimation.
  arXiv  Detail & Related papers  (2026-01-14T19:30:53Z)
• PHASE-Net: Physics-Grounded Harmonic Attention System for Efficient Remote Photoplethysmography Measurement [63.007237197267834]
  Existing deep learning methods are mostly physiological monitoring and lack theoretical robustness.<n>We propose a physics-informed r paradigm derived from the Navier-Stokes equations of hemodynamics, showing that the pulse signal follows a second-order system.<n>This provides a theoretical justification for using a Temporal Conal Network (TCN)<n>Phase-Net achieves state-of-the-art performance with strong efficiency, offering a theoretically grounded and deployment-ready r solution.
  arXiv  Detail & Related papers  (2025-09-29T14:36:45Z)
• Meta-training of diffractive meta-neural networks for super-resolution direction of arrival estimation [2.8163062697198744]
  We propose diffractive meta-neural networks (DMNNs) for accurate EM field modulation through metasurfaces.<n>DMNNs integrate pre-trained mini-metanets to characterize the amplitude and phase responses of meta-atoms across different polarizations and frequencies.<n>For wide-field super-resolution angle estimation, the system simultaneously resolves azimuthal and elevational angles through x and y-polarization channels.
  arXiv  Detail & Related papers  (2025-09-07T04:49:51Z)
• 1-bit Quantized On-chip Hybrid Diffraction Neural Network Enabled by Authentic All-optical Fully-connected Architecture [4.594367761345624]
  This study introduces the Hybrid Diffraction Neural Network (HDNN), a novel architecture that incorporates matrix multiplication into DNNs. utilizing a singular phase modulation layer and an amplitude modulation layer, the trained neural network demonstrated remarkable accuracies of 96.39% and 89% in digit recognition tasks.
  arXiv  Detail & Related papers  (2024-04-11T02:54:17Z)
• Classification robustness to common optical aberrations [64.08840063305313]
  This paper proposes OpticsBench, a benchmark for investigating robustness to realistic, practically relevant optical blur effects. Experiments on ImageNet show that for a variety of different pre-trained DNNs, the performance varies strongly compared to disk-shaped kernels. We show on ImageNet-100 with OpticsAugment that can be increased by using optical kernels as data augmentation.
  arXiv  Detail & Related papers  (2023-08-29T08:36:00Z)
• Physics-aware Roughness Optimization for Diffractive Optical Neural Networks [15.397285424104469]
  diffractive optical neural networks (DONNs) have shown promising advantages over conventional deep neural networks. We propose a physics-aware diffractive optical neural network training framework to reduce the performance difference between numerical modeling and practical deployment.
  arXiv  Detail & Related papers  (2023-04-04T03:19:36Z)
• Bayesian Neural Network Language Modeling for Speech Recognition [59.681758762712754]
  State-of-the-art neural network language models (NNLMs) represented by long short term memory recurrent neural networks (LSTM-RNNs) and Transformers are becoming highly complex. In this paper, an overarching full Bayesian learning framework is proposed to account for the underlying uncertainty in LSTM-RNN and Transformer LMs.
  arXiv  Detail & Related papers  (2022-08-28T17:50:19Z)
• Optical Phase Dropout in Diffractive Deep Neural Network [0.0]
  Unitary learning serves to update unitary weights update in deep complex-valued neural network with full connections. The square matrix property of unitary weights induces that the function signal has a limited dimension that could not generalize well. Phase dropout in unitary space that is evolved from a complex dropout and has a statistical inference is formulated for the first time. The equivalence between digital and diffractive model determines compound modulations that could successfully circumvent the nonlinear activations physically implemented in [DN]2.
  arXiv  Detail & Related papers  (2020-11-28T13:33:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.