Abstract:To provide a lightweight and cost-effective solution for the long-wave infrared imaging using a singlet, we develop a camera by integrating a High-Frequency-Enhancing Cycle-GAN neural network into a metalens imaging system. The High-Frequency-Enhancing Cycle-GAN improves the quality of the original metalens images by addressing inherent frequency loss introduced by the metalens. In addition to the bidirectional cyclic generative adversarial network, it incorporates a high-frequency adversarial learning module. This module utilizes wavelet transform to extract high-frequency components, and then establishes a high-frequency feedback loop. It enables the generator to enhance the camera outputs by integrating adversarial feedback from the high-frequency discriminator. This ensures that the generator adheres to the constraints imposed by the high-frequency adversarial loss, thereby effectively recovering the camera's frequency loss. This recovery guarantees high-fidelity image output from the camera, facilitating smooth video production. Our camera is capable of achieving dynamic imaging at 125 frames per second with an End Point Error value of 12.58. We also achieve 0.42 for Fréchet Inception Distance, 30.62 for Peak Signal to Noise Ratio, and 0.69 for Structural Similarity in the recorded videos.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to achieve a lightweight and cost - effective solution for high - quality, dynamic imaging in the long - wave infrared spectrum. Specifically, the paper proposes a metalens camera system based on the High - Frequency - Enhancing Cycle - GAN (HFE Cycle - GAN), aiming to improve the problem of image quality degradation caused by chromatic aberration and other optical aberrations during the imaging process of traditional metalenses. By integrating neural network technology, this system can effectively restore the high - frequency details in the metalens images, thereby improving the sharpness and detail performance of the images, enabling the camera to achieve high - quality video recording while remaining compact and lightweight. The key innovation points of the paper are as follows: 1. **High - frequency enhancement**: By introducing a high - frequency adversarial learning module, using wavelet transform to extract high - frequency components, and establishing a high - frequency feedback loop, the generator can enhance the camera output by integrating the adversarial feedback from the high - frequency discriminator, ensuring that the generator complies with the constraints of the high - frequency adversarial loss, thereby effectively restoring the frequency loss of the camera. 2. **Dynamic imaging ability**: This camera can achieve dynamic imaging at a speed of 125 frames per second, and has achieved good End Point Error values (12.58), Fréchet Inception Distance (0.42), Peak Signal to Noise Ratio (30.62) and Structural Similarity (0.69) in the recorded videos, indicating its excellent performance in video quality and smoothness. 3. **Applicability and compatibility**: This system is not only applicable to laboratory environments, but also compatible with commercial infrared cameras, expanding its potential in practical applications. In conclusion, this research aims to provide a new method for efficient, high - quality dynamic imaging in the long - wave infrared spectrum by combining advanced neural network technology and optimized metalens design.

Neural-Network-Enhanced Metalens Camera for High-Definition, Dynamic Imaging in the Long-Wave Infrared Spectrum

High-performance full-color imaging system based on end-to-end joint optimization of computer-generated holography and metalens.

Metalens-integrated Compact Imaging Devices for Wide-Field Microscopy

High-fidelity achromatic metalens imaging via deep neural network

Continuous Digital Zooming Using Generative Adversarial Networks for Dual Camera System

High-Resolution Metalens Imaging with Sequential Artificial Intelligence Models

Restoration of infrared metalens images with deep learning

Deep-learning-driven end-to-end metalens imaging

All-day Thin-Lens Computational Imaging with Scene-Specific Learning Recovery

Neural-Optic Co-Designed Polarization-Multiplexed Metalens for Compact Computational Spectral Imaging

High Efficiency Visible Achromatic Metalens Design via Deep Learning

Broadband Achromatic Polarization Insensitive Metalens Design in the Near-Infrared Spectrum

Nonlinear Computational Edge Detection Metalens

Neural nano-optics for high-quality thin lens imaging

Trilobite-inspired neural nanophotonic light-field camera with extreme depth-of-field

Synthetic Aperture Metalens

Light Field Imaging and Sensing Using a Broadband Achromatic Metalens Array

Feature-enhanced X-ray imaging using fused neural network strategy with designable metasurface

GAN-Based Focusing-Enhancement Method for Monochromatic Synthetic Aperture Imaging

Aberration-corrected hybrid metalens for longwave infrared thermal imaging

Planar wide-angle-imaging camera enabled by metalens array