Lensless speckle reconstructive spectrometer via physics-aware neural network

TL;DR

This paper introduces PhyspeNet, a physics-aware neural network that reconstructs spectra from speckle patterns without pre-training, enabling high-resolution, generalizable, and dataset-free spectral measurements using a lensless, snapshot system.

Contribution

The paper presents PhyspeNet, a novel untrained neural network framework combining CNN and physical models for speckle-based spectral reconstruction, eliminating the need for large datasets and training.

Findings

Distinguished dual-wavelength peaks separated by 2 pm.

Achieved a maximum bandwidth of 40 nm with high accuracy.

Demonstrated a lensless, snapshot spectrometer system.

Abstract

The speckle field yielded by disordered media is extensively employed for spectral measurements. Existing speckle reconstructive spectrometers (RSs) implemented by neural networks primarily rely on supervised learning, which necessitates large-scale spectra-speckle pairs. However, beyond system stability requirements for prolonged data collection, generating diverse spectra with high resolution and finely labeling them is particularly difficult. A lack of variety in datasets hinders the generalization of neural networks to new spectrum types. Here we avoid this limitation by introducing PhyspeNet, an untrained spectrum reconstruction framework combining a convolutional neural network (CNN) with a physical model of a chaotic optical cavity. Without pre-training and prior knowledge about the spectrum under test, PhyspeNet requires only a single captured speckle for various multi-wavelength reconstruction tasks. Experimentally, we demonstrate a lens-free, snapshot RS system by leveraging the one-to-many mapping between spatial and spectrum domains in a random medium. Dual-wavelength peaks separated by 2 pm can be distinguished, and a maximum working bandwidth of 40 nm is achieved with high measurement accuracy. This approach establishes a new paradigm for neural network-based RS systems, entirely eliminating reliance on datasets while ensuring that computational results exhibit a high degree of generalizability and physical explainability.