Searcharxiv⌕ Search

arXiv subjects

Yahui Zhang

Publications and source records attributed to Yahui Zhang.

18 recordsLinked to original sources
arXiv2026

CFRNet: Cycle-Consistent Fixed-Point Training for Real-Time Blind Face Restoration on Consumer Embedded NPUs

Blind face restoration on consumer devices has to balance image quality against speed and memory. Strong methods such as GFPGAN and CodeFormer give good perceptual quality, but they rely on large pretrained generative priors and on operators such as attention, codebook lookup, and style modulation that are hard to compile and quantize on the small neural processing units (NPUs) used in consumer hardware. Small convolutional restorers run fast enough, but they tend to over-smooth and to leave artifacts around the eyes, nose, and mouth. We present CFRNet, a 2.0,M-parameter ResNet-style restorer for on-device use at $256\times256$, the common face-crop size on consumer NPUs. The main idea is Cycle-Consistent Fixed-Point Training (CCFP). Instead of training the network for one pass and then running it several times by hand, we train it to act as a fixed-point operator, so that applying it again to a restored face does not change the face. CCFP uses three training losses, namely progressive multi-cycle supervision, an idempotence loss, and a re-degradation cycle loss, and it adds no cost at inference. To compare fairly under our deployment limits, we retrain all baselines from scratch at the same $256\times256$ resolution. On a 300-image test set, CFRNet reaches the best perceptual score (LPIPS 0.250 at three cycles, which is 31% lower than one cycle) and also the best PSNR and SSIM at two cycles. It runs in about 23,ms per cycle in INT8 on a HiSilicon Hi3402 NPU, while the same baselines cannot be compiled to that chip. The cycle count $k$ acts as a simple quality knob that needs no retraining: PSNR is best at $k\!=\!2$ and LPIPS keeps improving up to $k\!=\!3$. We further show that the same idea works with a plain CNN that is even easier to deploy, and we run the model in real time on an in-car driver-monitoring board.

cs.CV↗
arXiv2026

Hardware Implementation of Photonic Spiking Hash Retrieval

Hashing retrieval is a pivotal technology for large-scale similarity search, widely applied in retrieval-augmented generation (RAG) for large language models (LLMs), massive image repositories, and bioinformatics sequence matching. However, traditional electronic hashing implementations face severe bottlenecks in power consumption and latency when processing high-dimensional data, while existing photonic neural networks often lack robust mechanisms for direct binary code generation under analog noise. To address these challenges, we propose a hardware-software co-designed photonic spiking hashing framework. We utilize the nonlinear thresholding dynamics of a distributed feedback laser with saturable absorber (DFB-SA) to realize the final binarization of a single-step spiking neural network (SNN). Crucially, a hardware-aware quantization margin loss is introduced to maximize the decision margin, effectively mitigating bit flips caused by optical intensity fluctuations. Validated on MNIST (image) and 20 Newsgroups (text) datasets, our system demonstrates robust binary code generation and high retrieval accuracy comparable to digital baselines. Most significantly, the proposed photonic architecture exhibits superior efficiency with an encoding latency of 2.294 ns/query and an energy consumption of 73.70 pJ/query. This work offers a robust and viable path for ultra-fast, energy-efficient optoelectronic neuromorphic computing in high-throughput information retrieval tasks.

physics.optics↗
arXiv2026

Hardware implementation of photonic neuromorphic autonomous navigation

Reinforcement learning (RL) is a core technology enabling the transition of artificial intelligence (AI) from perception to decision-making, but its deployment on conventional electronic hardware suffers from high latency and energy consumption imposed by the von Neumann architecture. Here, we propose a photonic spiking twin delayed deep deterministic policy gradient (TD3) reinforcement learning architecture for neuromorphic autonomous navigation and experimentally validate it on a distributed feedback laser with a saturable absorber (DFB-SA) array. The hybrid architecture integrates a photonic spiking Actor network with dual continuous-valued Critic networks, where the final nonlinear spiking activation layer of the Actor is deployed on the DFB-SA laser array. In autonomous navigation tasks, the system achieves an average reward of 58.22 plus-minus 17.29 and a success rate of 80% plus-minus 8.3%. Hardware-software co-inference demonstrates an estimated energy consumption of 0.78 nJ/inf and an ultra-low latency of 191.20 ps/inf, with co-inference error rates of 0.051% and 0.059% in task scenarios with and without obstacle interference, respectively. Simulations for error-activated channels show full agreement with the expected responses, validating the dynamic characteristics of the DFB-SA laser. The architecture shows strong potential for integration with large-scale photonic linear computing chips, enabling fully-functional photonic computation and low-power, low-latency neuromorphic autonomous navigation.

physics.optics↗
arXiv2026

Photonic spiking reinforcement learning for intelligent routing

Intelligent routing plays a key role in modern communication infrastructure, including data centers, computing networks, and future 6G networks. Although reinforcement learning (RL) has shown great potential for intelligent routing, its practical deployment remains constrained by high energy consumption and decision latency. Here, we propose a photonic spiking RL architecture that implements a proximal policy optimization (PPO)-based intelligent routing algorithm. The performance of the proposed approach is systematically evaluated on a software-defined network (SDN) with a fat-tree topology. The results demonstrate that, under various baseline traffic rate conditions, the PPO-based routing strategy significantly outperforms the conventional Dijkstra algorithm in several key performance metrics. Furthermore, a hardware-software collaborative framework of the spiking Actor network is realized for three typical baseline traffic rates, utilizing a photonic synapse chip based on a Mach-Zehnder interferometer (MZI) array and a photonic spiking neuron chip based on distributed feedback lasers with a saturable absorber (DFB-SAs). Experimental validation on 640 state-action pairs shows that the inference accuracy of the hardware-software collaborative framework is consistent with that of the pure algorithmic implementation. The impacts of different hidden-layer scales in the spiking Actor network and varying network size of fat-tree topology are further analyzed. The integration of photonic spiking RL with SDN-based routing establishes a novel paradigm for intelligent routing optimization, featuring ultra-low latency and high energy efficiency. This approach exhibits broad application prospects in real-time network optimization scenarios, including large-scale data centers, computing networks, satellite Internet systems, and future 6G networks.

physics.optics↗
arXiv2025

Photonic Spiking Graph Neural Network for Energy-Efficient Structured Data Processing

Photonic computing shows great potential for signal processing and artificial intelligence (AI) acceleration due to its ultra-high speed, low energy consumption, and inherent parallelism. Existing photonic computing research has mainly focused on convolutional neural networks (CNNs) and fully connected neural networks (FCNNs), which are well suited for tasks such as image classification and object detection but face limitations in handling graph-structured data. Graph neural networks (GNNs) are specifically designed to model complex relational structures. In this work, we propose a photonic spiking graph neural network (PSGNN) architecture that integrates the structural modeling capability of GNNs, the temporal dynamics of spiking neurons, and the parallel computing advantages of photonic hardware. Through hardware-software co-optimization, a bias-term simulation method tailored for photonic chips is implemented using feature-dimension expansion, enabling effective network training. Experiments on the KarateClub and PubMed datasets achieve training accuracies of 100 percent (92 +/- 2 percent) and test accuracies of 97 percent (90 +/- 1 percent). A silicon photonics 4 x 4 Mach-Zehnder interferometer (MZI) array is further constructed for hardware validation, achieving a test accuracy of 93 percent. The system demonstrates an inference latency of 97 ps, with an energy efficiency of 280 GOPS/W and a computational density of 52 GOPS/mm^2. These results highlight the potential of PSGNN for structured-data processing applications.

physics.optics↗
arXiv2025

Bayes-DIC Net: Estimating Digital Image Correlation Uncertainty with Bayesian Neural Networks

This paper introduces a novel method for generating high-quality Digital Image Correlation (DIC) dataset based on non-uniform B-spline surfaces. By randomly generating control point coordinates, we construct displacement fields that encompass a variety of realistic displacement scenarios, which are subsequently used to generate speckle pattern datasets. This approach enables the generation of a large-scale dataset that capture real-world displacement field situations, thereby enhancing the training and generalization capabilities of deep learning-based DIC algorithms. Additionally, we propose a novel network architecture, termed Bayes-DIC Net, which extracts information at multiple levels during the down-sampling phase and facilitates the aggregation of information across various levels through a single skip connection during the up-sampling phase. Bayes-DIC Net incorporates a series of lightweight convolutional blocks designed to expand the receptive field and capture rich contextual information while minimizing computational costs. Furthermore, by integrating appropriate dropout modules into Bayes-DIC Net and activating them during the network inference stage, Bayes-DIC Net is transformed into a Bayesian neural network. This transformation allows the network to provide not only predictive results but also confidence levels in these predictions when processing real unlabeled datasets. This feature significantly enhances the practicality and reliability of our network in real-world displacement field prediction tasks. Through these innovations, this paper offers new perspectives and methods for dataset generation and algorithm performance enhancement in the field of DIC.

cs.CV↗
arXiv2025

Hardware-aware Lightweight Photonic Spiking Neural Network for Pattern Classification

There exists a significant scale gap between photonic neural network integrated chips and neural networks, which hinders the deployment and application of photonic neural network. Here, we propose hardware-aware lightweight spiking neural networks (SNNs) architecture tailored to our photonic neuromorphic chips, and conducts hardware-software collaborative computing for solving patter classification tasks. Here, we employed a simplified Mach-Zehnder interferometer (MZI) mesh for performing linear computation, and 16-channel distributed feedback lasers with saturable absorber (DFB-SA) array for performing nonlinear spike activation. Both photonic neuromorphic chips based on the MZI mesh and DFB-SA array were designed, optimized and fabricated. Furthermore, we propose a lightweight spiking neural network (SNN) with discrete cosine transform to reduce input dimension and match the input/output ports number of the photonic neuromorphic chips. We demonstrated an end-to-end inference of an entire layer of the lightweight photonic SNN. The hardware-software collaborative inference accuracy is 90% and 80.5% for MNIST and Fashion-MNIST datasets, respectively. The energy efficiency is 1.39 TOPS/W for the MZI mesh, and is 987.65 GOPS/W for the DFB-SA array. The lightweight architecture and experimental demonstration address the challenge of scale mismatch between the photonic chip and SNN, paving the way for the hardware deployment of photonic SNNs.

physics.optics↗
arXiv2025

Hardware-Software Collaborative Computing of Photonic Spiking Reinforcement Learning for Robotic Continuous Control

Robotic continuous control tasks impose stringent demands on the energy efficiency and latency of computing architectures due to their high-dimensional state spaces and real-time interaction requirements. Conventional electronic computing platforms face computational bottlenecks, whereas the fusion of photonic computing and spiking reinforcement learning (RL) offers a promising alternative. Here, we propose a novel computing architecture based on photonic spiking RL, which integrates the Twin Delayed Deep Deterministic policy gradient (TD3) algorithm with spiking neural network (SNN). The proposed architecture employs an optical-electronic hybrid computing paradigm wherein a silicon photonic Mach-Zehnder interferometer (MZI) chip executes linear matrix computations, while nonlinear spiking activations are performed in the electronic domain. Experimental validation on the Pendulum-v1 and HalfCheetah-v2 benchmarks demonstrates the system capability for software-hardware co-inference, achieving a control policy reward of 5831 on HalfCheetah-v2, a 23.33% reduction in convergence steps, and an action deviation below 2.2%. Notably, this work represents the first application of a programmable MZI photonic computing chip to robotic continuous control tasks, attaining an energy efficiency of 1.39 TOPS/W and an ultralow computational latency of 120 ps. Such performance underscores the promise of photonic spiking RL for real-time decision-making in autonomous and industrial robotic systems.

cs.RO↗
arXiv2025

Integrated photonic neuromorphic computing: device, architecture, chip, algorithm

Artificial intelligence (AI) has experienced explosive growth in recent years. The large models have been widely applied in various fields, including natural language processing, image generation, and complex decision-making systems, revolutionizing technological paradigms across multiple industries. Nevertheless, the substantial data processing demands during model training and inference result in the computing power bottleneck. Traditional electronic chips based on the von Neumann architecture struggle to meet the growing demands for computing power and power efficiency amid the continuous development of AI. Photonic neuromorphic computing, an emerging solution in the post-Moore era, exhibits significant development potential. Leveraging the high-speed and large-bandwidth characteristics of photons in signal transmission, as well as the low-power consumption advantages of optical devices, photonic integrated computing chips have the potential to overcome the memory wall and power wall issues of electronic chips. In recent years, remarkable advancements have been made in photonic neuromorphic computing. This article presents a systematic review of the latest research achievements. It focuses on fundamental principles and novel neuromorphic photonic devices, such as photonic neurons and photonic synapses. Additionally, it comprehensively summarizes the network architectures and photonic integrated neuromorphic chips, as well as the optimization algorithms of photonic neural networks. In addition, combining with the current status and challenges of this field, this article conducts an in-depth discussion on the future development trends of photonic neuromorphic computing in the directions of device integration, algorithm collaborative optimization, and application scenario expansion, providing a reference for subsequent research in the field of photonic neuromorphic computing.

physics.optics↗
arXiv2025

Nonlinear Photonic Neuromorphic Chips for Spiking Reinforcement Learning

Photonic computing chips have made significant progress in accelerating linear computations, but nonlinear computations are usually implemented in the digital domain, which introduces additional system latency and power consumption, and hinders the implementation of fully-functional photonic neural network chips. Here, we propose and fabricate a 16-channel programmable incoherent photonic neuromorphic computing chip by co-designing a simplified MZI mesh and distributed feedback lasers with saturable absorber array using different materials, enabling implementation of both linear and nonlinear spike computations in the optical domain. Furthermore, previous studies mainly focused on supervised learning and simple image classification tasks. Here, we propose a photonic spiking reinforcement learning (RL) architecture for the first time, and develop a software-hardware collaborative training-inference framework to address the challenge of training spiking RL models. We achieve large-scale, energy-efficient (photonic linear computation: 1.39 TOPS/W, photonic nonlinear computation: 987.65 GOPS/W) and low-latency (320 ps) end-to-end deployment of an entire layer of photonic spiking RL. Two RL benchmarks include the discrete CartPole task and the continuous Pendulum tasks are demonstrated experimentally based on spiking proximal policy optimization algorithm. The hardware-software collaborative computing reward value converges to 200 (-250) for the CartPole tasks, respectively, comparable to that of a traditional PPO algorithm. This experimental demonstration addresses the challenge of the absence of large-scale photonic nonlinear spike computation and spiking RL training difficulty, and presents a high-speed and low-latency photonic spiking RL solution with promising application prospects in fields such as real-time decision-making and control for robots and autonomous driving.

physics.optics↗
arXiv2023

Two-dimension to three-dimension transition of chiral spin liquid and fractional quantum Hall phases

There have been lots of interest in two-dimensional (2D) fractional phases with an emergent U (1) gauge field. However, many experimental realizations are actually in three-dimensional (3D) systems with infinitely stacked 2D layers. Then a natural question arises: starting from the decoupling limit with 2+1d U (1) gauge field in each layer, how does the gauge field become 3+1d when increasing inter-layer coupling? Here we propose a 2D to 3D transition through condensing inter-layer exciton. The Goldstone mode of the condensation becomes the missing az component in the 3D phase. As a simple example, we construct a 3D chiral spin liquid (CSL) from infinitely stacked 2D CSL. The 3D CSL has a gapless photon mode with dispersion $ω\sim q_z^2$ in the z-direction. The same theory also applies to the fractional quantum Hall phase. At the 2D to the 3D transition point, there are gapless modes at each $q_z$ along a line $q = (0,0,q_z)$ in momentum space, in contrast to a conventional critical point with gapless mode only at one momentum. Meanwhile, the scaling dimension $Δ(q_z)$ has $q_z$ dependence, indicating a more non-trivial structure than a simple decoupled fixed point. Our theory can also be generalized to a critical point between a generic infinite component Chern-Simons-Maxwell theory (iCSM) with both intra-layer and inter-layer Chern Simons term and a 3D gapless phase. Certain iCSM theories have recently been shown to describe gapped non-foliated fracton orders. Therefore we have a continuous transition between a gapped fracton order and a 3D gapless phase.

cond-mat.str-el↗
arXiv2023

Photonic Integrated Neuro-Synaptic Core for Convolutional Spiking Neural Network

Neuromorphic photonic computing has emerged as a competitive computing paradigm to overcome the bottlenecks of the von-Neumann architecture. Linear weighting and nonlinear spiking activation are two fundamental functions of a photonic spiking neural network (PSNN). However, they are separately implemented with different photonic materials and devices, hindering the large-scale integration of PSNN. Here, we propose, fabricate and experimentally demonstrate a photonic neuro-synaptic chip enabling the simultaneous implementation of linear weighting and nonlinear spiking activation based on a distributed feedback (DFB) laser with a saturable absorber (DFB-SA). A prototypical system is experimentally constructed to demonstrate the parallel weighted function and nonlinear spike activation. Furthermore, a four-channel DFB-SA array is fabricated for realizing matrix convolution of a spiking convolutional neural network, achieving a recognition accuracy of 87% for the MNIST dataset. The fabricated neuro-synaptic chip offers a fundamental building block to construct the large-scale integrated PSNN chip.

physics.optics↗
arXiv2022

Spiking SiamFC++: Deep Spiking Neural Network for Object Tracking

Spiking neural network (SNN) is a biologically-plausible model and exhibits advantages of high computational capability and low power consumption. While the training of deep SNN is still an open problem, which limits the real-world applications of deep SNN. Here we propose a deep SNN architecture named Spiking SiamFC++ for object tracking with end-to-end direct training. Specifically, the AlexNet network is extended in the time domain to extract the feature, and the surrogate gradient function is adopted to realize direct supervised training of the deep SNN. To examine the performance of the Spiking SiamFC++, several tracking benchmarks including OTB2013, OTB2015, VOT2015, VOT2016, and UAV123 are considered. It is found that, the precision loss is small compared with the original SiamFC++. Compared with the existing SNN-based target tracker, e.g., the SiamSNN, the precision (succession) of the proposed Spiking SiamFC++ reaches 85.24% (64.37%), which is much higher than that of 52.78% (44.32%) achieved by the SiamSNN. To our best knowledge, the performance of the Spiking SiamFC++ outperforms the existing state-of-the-art approaches in SNN-based object tracking, which provides a novel path for SNN application in the field of target tracking. This work may further promote the development of SNN algorithms and neuromorphic chips.

cs.CV↗
arXiv2022

Photonic p-orbital higher-order topological insulators

The orbital degrees of freedom play a pivotal role in understanding fundamental phenomena in solid-state materials as well as exotic quantum states of matter including orbital superfluidity and topological semimetals. Despite tremendous efforts in engineering synthetic cold-atom, electronic and photonic lattices to explore orbital physics, thus far high orbitals in an important class of materials, namely, the higher-order topological insulators (HOTIs), have not been realized. Here, we demonstrate p-orbital corner states in a photonic HOTI, unveiling their underlying topological invariant, symmetry protection, and nonlinearity-induced dynamical rotation. In a Kagome-type HOTI, we find that topological protection of the p-orbital corner states demands an orbital-hopping symmetry, in addition to the generalized chiral symmetry. Due to orbital hybridization, the nontrivial topology of the p-orbital HOTI is hidden if bulk polarization is used as the topological invariant, but well manifested by the generalized winding number. Our work opens a pathway for the exploration of intriguing orbital phenomena mediated by higher band topology applicable to a broad spectrum of systems.

physics.optics↗
arXiv2022

Hardware-algorithm collaborative computing with photonic spiking neuron chip based on integrated Fabry-Pérot laser with saturable absorber

Photonic neuromorphic computing has emerged as a promising avenue toward building a low-latency and energy-efficient non-von-Neuman computing system. Photonic spiking neural network (PSNN) exploits brain-like spatiotemporal processing to realize high-performance neuromorphic computing. However, the nonlinear computation of PSNN remains a significant challenging. Here, we proposed and fabricated a photonic spiking neuron chip based on an integrated Fabry-Pérot laser with a saturable absorber (FP-SA) for the first time. The nonlinear neuron-like dynamics including temporal integration, threshold and spike generation, refractory period, and cascadability were experimentally demonstrated, which offers an indispensable fundamental building block to construct the PSNN hardware. Furthermore, we proposed time-multiplexed spike encoding to realize functional PSNN far beyond the hardware integration scale limit. PSNNs with single/cascaded photonic spiking neurons were experimentally demonstrated to realize hardware-algorithm collaborative computing, showing capability in performing classification tasks with supervised learning algorithm, which paves the way for multi-layer PSNN for solving complex tasks.

cs.ET↗
arXiv2020

All-optical neuromorphic binary convolution with a spiking VCSEL neuron for image gradient magnitudes

All-optical binary convolution with a photonic spiking vertical-cavity surface-emitting laser (VCSEL) neuron is proposed and demonstrated experimentally for the first time. Optical inputs, extracted from digital images and temporally encoded using rectangular pulses, are injected in the VCSEL neuron which delivers the convolution result in the number of fast (<100 ps long) spikes fired. Experimental and numerical results show that binary convolution is achieved successfully with a single spiking VCSEL neuron and that all-optical binary convolution can be used to calculate image gradient magnitudes to detect edge features and separate vertical and horizontal components in source images. We also show that this all-optical spiking binary convolution system is robust to noise and can operate with high-resolution images. Additionally, the proposed system offers important advantages such as ultrafast speed, high energy efficiency and simple hardware implementation, highlighting the potentials of spiking photonic VCSEL neurons for high-speed neuromorphic image processing systems and future photonic spiking convolutional neural networks.

physics.optics↗
arXiv2020

Convolutional Image Edge Detection Using Ultrafast Photonic Spiking VCSEL Neurons

We report experimentally and in theory on the detection of edge information in digital images using ultrafast spiking optical artificial neurons towards convolutional neural networks (CNNs). In tandem with traditional convolution techniques, a photonic neuron model based on a Vertical-Cavity Surface Emitting Laser (VCSEL) is implemented experimentally to threshold and activate fast spiking responses upon the detection of target edge features in digital images. Edges of different directionalities are detected using individual kernel operators and complete image edge detection is achieved using gradient magnitude. Importantly, the neuromorphic (brain-like) image edge detection system of this work uses commercially sourced VCSELs exhibiting spiking responses at sub-nanosecond rates (many orders of magnitude faster than biological neurons) and operating at the telecom wavelength of 1300 nm; hence making our approach compatible with optical communication and data-center technologies. These results therefore have exciting prospects for ultrafast photonic implementations of neural networks towards computer vision and decision making systems for future artificial intelligence applications.

eess.IV↗