arXiv subjects
Publications and source records attributed to Niyazi Ulas Dinc.
This roadmap consolidates recent advances while exploring emerging applications, reflecting the remarkable diversity of hardware platforms, neuromorphic concepts, and implementation philosophies reported in the field. It emphasizes the critical role of cross-disciplinary collaboration in this rapidly evolving field.
The ability to train ever-larger neural networks brings artificial intelligence to the forefront of scientific and technical discoveries. However, their exponentially increasing size creates a proportionally greater demand for energy and computational hardware. Incorporating complex physical events in networks as fixed, efficient computation modules can address this demand by decreasing the complexity of trainable layers. Here, we utilize ultrashort pulse propagation in multimode fibers, which perform large-scale nonlinear transformations, for this purpose. Training the hybrid architecture is achieved through a neural model that differentiably approximates the optical system. The training algorithm updates the neural simulator and backpropagates the error signal over this proxy to optimize layers preceding the optical one. Our experimental results achieve state-of-the-art image classification accuracies and simulation fidelity. Moreover, the framework demonstrates exceptional resilience to experimental drifts. By integrating low-energy physical systems into neural networks, this approach enables scalable, energy-efficient AI models with significantly reduced computational demands.
Diffusion models generate new samples by progressively decreasing the noise from the initially provided random distribution. This inference procedure generally utilizes a trained neural network numerous times to obtain the final output, creating significant latency and energy consumption on digital electronic hardware such as GPUs. In this study, we demonstrate that the propagation of a light beam through a semi-transparent medium can be programmed to implement a denoising diffusion model on image samples. This framework projects noisy image patterns through passive diffractive optical layers, which collectively only transmit the predicted noise term in the image. The optical transparent layers, which are trained with an online training approach, backpropagating the error to the analytical model of the system, are passive and kept the same across different steps of denoising. Hence this method enables high-speed image generation with minimal power consumption, benefiting from the bandwidth and energy efficiency of optical information processing.
Electronic computers have evolved drastically over the past years with an ever-growing demand for improved performance. However, the transfer of information from memory and high energy consumption have emerged as issues that require solutions. Optical techniques are considered promising solutions to these problems with higher speed than their electronic counterparts and with reduced energy consumption. Here, we use the optical reservoir computing framework we have previously described (Scalable Optical Learning Operator or SOLO) to program the spatial-spectral output of the light after nonlinear propagation in a multimode fiber. The novelty in the current paper is that the system is programmed through an output sampling scheme, similar to that used in hyperspectral imaging in astronomy. Linear and nonlinear computations are performed by light in the multimode fiber and the high dimensional spatial-spectral information at the fiber output is optically programmed before it reaches the camera. We then used a digital computer to classify the programmed output of the multi-mode fiber using a simple, single layer network. When combining front-end programming and the proposed spatial-spectral programming, we were able to achieve 89.9% classification accuracy on the dataset consisting of chest X-ray images from COVID-19 patients. At the same time, we obtained a decrease of 99% in the number of tunable parameters compared to an equivalently performing digital neural network. These results show that the performance of programmed SOLO is comparable with cutting-edge electronic computing platforms, albeit with a much-reduced number of electronic operations.
Artificial Intelligence (AI) demands large data flows within datacenters, heavily relying on multicasting data transfers. As AI models scale, the requirement for high-bandwidth and low-latency networking compounds. The common use of electrical packet switching faces limitations due to optical-electrical-optical conversion bottlenecks. Optical switches, while bandwidth-agnostic and low-latency, suffer from having only unicast or non-scalable multicasting capability. This paper introduces an optical switching technique addressing this challenge. Our approach enables arbitrarily programmable simultaneous unicast and multicast connectivity, eliminating the need for optical splitters that hinder scalability due to optical power loss. We use phase modulation in multiple layers, tailored to implement any multicast connectivity map. Phase modulation also enables wavelength selectivity on top of spatial selectivity, resulting in an optical switch that implements space-wavelength routing. We conducted simulations and experiments to validate our approach. Our results affirm the concept's feasibility, effectiveness, and scalability, as a multicasting switch by experimentally demonstrating 16 spatial ports using 2 wavelength channels. Numerically, 64 spatial ports with 4 wavelength channels each were simulated, with approximately constant efficiency (< 3 dB) as ports and wavelength channels scale.
Deep neural networks have achieved remarkable breakthroughs by leveraging multiple layers of data processing to extract hidden representations, albeit at the cost of large electronic computing power. To enhance energy efficiency and speed, the optical implementation of neural networks aims to harness the advantages of optical bandwidth and the energy efficiency of optical interconnections. In the absence of low-power optical nonlinearities, the challenge in the implementation of multilayer optical networks lies in realizing multiple optical layers without resorting to electronic components. In this study, we present a novel framework that uses multiple scattering that is capable of synthesizing programmable linear and nonlinear transformations concurrently at low optical power by leveraging the nonlinear relationship between the scattering potential, represented by data, and the scattered field. Theoretical and experimental investigations show that repeating the data by multiple scattering enables non-linear optical computing at low power continuous wave light. Moreover, we empirically found that scaling of this optical framework follows the power law as in state-of-the-art deep digital networks.
Phase imaging is widely used in biomedical imaging, sensing, and material characterization, among other fields. However, direct imaging of phase objects with subwavelength resolution remains a challenge. Here, we demonstrate subwavelength imaging of phase and amplitude objects based on all-optical diffractive encoding and decoding. To resolve subwavelength features of an object, the diffractive imager uses a thin, high-index solid-immersion layer to transmit high-frequency information of the object to a spatially-optimized diffractive encoder, which converts/encodes high-frequency information of the input into low-frequency spatial modes for transmission through air. The subsequent diffractive decoder layers (in air) are jointly designed with the encoder using deep-learning-based optimization, and communicate with the encoder layer to create magnified images of input objects at its output, revealing subwavelength features that would otherwise be washed away due to diffraction limit. We demonstrate that this all-optical collaboration between a diffractive solid-immersion encoder and the following decoder layers in air can resolve subwavelength phase and amplitude features of input objects in a highly compact design. To experimentally demonstrate its proof-of-concept, we used terahertz radiation and developed a fabrication method for creating monolithic multi-layer diffractive processors. Through these monolithically fabricated diffractive encoder-decoder pairs, we demonstrated phase-to-intensity transformations and all-optically reconstructed subwavelength phase features of input objects by directly transforming them into magnified intensity features at the output. This solid-immersion-based diffractive imager, with its compact and cost-effective design, can find wide-ranging applications in bioimaging, endoscopy, sensing and materials characterization.
We demonstrate the fabrication of volume holograms using 2-photon polymerization with dynamic control of light exposure. We refer to our method as (3+1)D printing. Volume holograms that are recorded by interfering reference and signal beams have a diffraction efficiency relation that is inversely proportional with the square of the number of superimposed holograms. By using (3+1)D printing for fabrication, the refractive index of each voxel is created independently and thus by, digitally filtering the undesired interference terms, the diffraction efficiency is now inversely proportional to the number of multiplexed gratings. We experimentally demonstrated this linear dependence by recording M=50 volume gratings. To the best of our knowledge, this is the first experimental demonstration of distributed volume holograms that overcome the 1/M^2 limit.
Neural networks (NN) have demonstrated remarkable capabilities in various tasks, but their computation-intensive nature demands faster and more energy-efficient hardware implementations. Optics-based platforms, using technologies such as silicon photonics and spatial light modulators, offer promising avenues for achieving this goal. However, training multiple trainable layers in tandem with these physical systems poses challenges, as they are difficult to fully characterize and describe with differentiable functions, hindering the use of error backpropagation algorithm. The recently introduced Forward-Forward Algorithm (FFA) eliminates the need for perfect characterization of the learning system and shows promise for efficient training with large numbers of programmable parameters. The FFA does not require backpropagating an error signal to update the weights, rather the weights are updated by only sending information in one direction. The local loss function for each set of trainable weights enables low-power analog hardware implementations without resorting to metaheuristic algorithms or reinforcement learning. In this paper, we present an experiment utilizing multimode nonlinear wave propagation in an optical fiber demonstrating the feasibility of the FFA approach using an optical system. The results show that incorporating optical transforms in multilayer NN architectures trained with the FFA, can lead to performance improvements, even with a relatively small number of trainable weights. The proposed method offers a new path to the challenge of training optical NNs and provides insights into leveraging physical transformations for enhancing NN performance.
The ever-increasing demand for processing data with larger machine learning models requires more efficient hardware solutions due to limitations such as power dissipation and scalability. Optics is a promising contender for providing lower power computation since light propagation through a non-absorbing medium is a lossless operation. However, to carry out useful and efficient computations with light, generating and controlling nonlinearity optically is a necessity that is still elusive. Multimode fibers (MMF) have been shown that they can provide nonlinear effects with microwatts of average power while maintaining parallelism and low loss. In this work, we propose an optical neural network architecture, which performs nonlinear optical computation by controlling the propagation of ultrashort pulses in MMF by wavefront shaping. With a surrogate model, optimal sets of parameters are found to program this optical computer for different tasks with minimal utilization of an electronic computer. We show a remarkable decrease of 97% in the number of model parameters, which leads to an overall 99% digital operation reduction compared to an equivalently performing digital neural network. We further demonstrate that a fully optical implementation can also be performed with competitive accuracies.
We propose single-step additive fabrication of graded-index optical elements by introducing the light exposure as the additional dimension to three-dimensional (3D) laser writing, hence (3+1)D writing. We use a commercial printer and photoresist to realize the proposed single step fabrication method that can be swiftly adopted for research and engineering. After presenting the characterization of the graded-index profiles via basic structures, we demonstrate two different optical devices: volume holograms that are superimposed using angular and peristrophic multiplexing, and optical waveguides with well-defined refractive index profiles. In the latter, we precisely control the propagating modes via tuning the (3+1)D-printed waveguide parameters and report step-index and graded-index core-cladding transitions.
We motivate a new canonical strategy for integrating photonic neural networks (NNs) by leveraging 3D printing. Our believe is that a NN's parallel and dense connectivity is not scalable without 3D integration. 3D additive fabrication complemented with photonic signal transduction can dramatically augment the current capabilities of 2D CMOS and integrated photonics. Here we review some of our recent advances made towards such a breakthrough architecture.