Recently, free-space optical neural networks (ONNs) have gained extensive interest as emerging machine learning platforms for implementing artificial intelligence tasks, such as image classification. Despite various optical implementations of electronic neural networks (ENNs), the bulky volume of optical components remains challenging to deploy edge devices, such as Internet of Things peripherals, wearable devices, and camera. To address this problem, we propose a compact lensless optoelectronic convolutional neural network (LOE-CNN) architecture with a lensless optical analog processor utilizing a single optimized diffractive phase mask (DPM) to perform convolution operations without Fourier lens. Comparing the processor with a commercially available NVIDIA A100 Tensor Core GPU in terms of speed and power, indicates the optical computing platform enables to replace the electronic processor in latency reduction and energy savings. Furthermore, we compare the LOE-CNN with two all-electronic neural networks (i.e., fully connected neural network [FC-NN] and convolutional neural network [CNN]) over the Modified National Institute of Standards and Technology (MNIST) dataset and Fashion-MNIST dataset, respectively, and demonstrate that the LOE-CNN can be functionally comparable to existing electronic counterparts in classification performance. My study not only opens up new application prospects for free-space ONNs based on compact lensless single-chip convolution processor, but also facilitates the development of ONNs-based smart devices.
Traditional analytical algorithm needs to combine the transmission functions of grating and lens to generate a Computer Generated Hologram (CGH), so as to realize the distribution of three-dimensional (3D) multi-focal points in space, but the grating phase will inevitably produce high-order diffraction focus, resulting in energy loss, and the traditional analytic algorithm is more suitable for generating array multi-focal distribution with equal spacing. To solve this problem, this paper simplifies the traditional analytical algorithm, and proposes a method that only uses multi-lens phase and random phase superposition to generate the CGH required by the target light location, by changing the focal length of the lens phase, the multi-focus distribution along the z-axial direction of multiple independent focal planes is realized. Then the phase of these different focal planes is superimposed, and a 0~2π random phase modulation is added, which can quickly generate 3D multi-focus distribution with controllable number and position. The simulation results show that the energy uniformity of focal spot on each focal plane is between 89.45% and 98.08%. The experimental results show that the energy uniformity of focal spots on each focal plane is between 88.40% and 96.13%, which is consistent with the simulation results. Compared with traditional analytical algorithm, the proposed method is more universal for multi-focus distribution in 3D space without special requirements of array distribution with equal spacing, and has potential application value in laser processing, holographic optical tweezers, optical communication and other fields.
Shack-Hartmann sensor is widely used in adaptive optics systems, and laser beam quality measurements. The traditional method separates measures and calculations, and the wavefront reconstruction algorithm is slow to implement on the host computer. In this paper, the embedded GPU is introduced to Shack-Hartmann sensors' wavefront phase reconstruction. A parallel calculation method is proposed to speed up the wavefront phase reconstruction process. The experiment result shows the algorithm speed improves 50× with the image size of 2592×2048 pixels.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.