Optical Physics-Based Generative Models

This paper establishes a comprehensive mathematical framework connecting optical physics equations to generative models, demonstrating how light propagation dynamics inspire powerful artificial intelligence approaches. We analyze six fundamental optical equations, comparing linear models (Helmholtz, dissipative wave, and Eikonal equations) with their nonlinear extensions incorporating Kerr effects, cubic-quintic nonlinearities, and intensity-dependent refractive indices. Our nonlinear optical models reveal remarkable capabilities through natural self-organization principles. The nonlinear Helmholtz model achieves 40-60% parameter reduction while maintaining superior mode separation via self-focusing phenomena. The cubic-quintic dissipative wave model prevents mode collapse through balanced attractive-repulsive interactions, enabling stable soliton formation with 20-40% improved coverage. The intensity-dependent Eikonal model creates adaptive pathways that dynamically respond to content, providing enhanced controllability in conditional generation. Experimental validation demonstrates consistent superiority over linear predecessors and traditional generative approaches. The nonlinear Helmholtz model achieves FID scores of 0.0089 versus 1.0909 for linear versions, while the cubic-quintic model reaches 0.0156 FID with exceptional stability. Memory usage drops 40-60% and training time improves 30-50% due to inherent nonlinear stability properties. The framework enables bidirectional benefits, advancing both generative AI and optical physics through novel approaches to soliton analysis, wavefront control, and refractive index reconstruction with 95% accuracy. This work reveals deep connections between physical self-organization and artificial intelligence, opening pathways toward efficient optical computing implementations.

Training Large-Scale Optical Neural Networks with Two-Pass Forward Propagation

This paper addresses the limitations in Optical Neural Networks (ONNs) related to training efficiency, nonlinear function implementation, and large input data processing. We introduce Two-Pass Forward Propagation, a novel training method that avoids specific nonlinear activation functions by modulating and re-entering error with random noise. Additionally, we propose a new way to implement convolutional neural networks using simple neural networks in integrated optical systems. Theoretical foundations and numerical results demonstrate significant improvements in training speed, energy efficiency, and scalability, advancing the potential of optical computing for complex data tasks.

Free-Space Optical Spiking Neural Network

Neuromorphic engineering has emerged as a promising avenue for developing brain-inspired computational systems. However, conventional electronic AI-based processors often encounter challenges related to processing speed and thermal dissipation. As an alternative, optical implementations of such processors have been proposed, capitalizing on the intrinsic information-processing capabilities of light. Within the realm of optical neuromorphic engineering, various optical neural networks (ONNs) have been explored. Among these, Spiking Neural Networks (SNNs) have exhibited notable success in emulating the computational principles of the human brain. Nevertheless, the integration of optical SNN processors has presented formidable obstacles, mainly when dealing with the computational demands of large datasets. In response to these challenges, we introduce a pioneering concept: the Free-space Optical deep Spiking Convolutional Neural Network (OSCNN). This novel approach draws inspiration from computational models of the human eye. We have meticulously designed various optical components within the OSCNN to tackle object detection tasks across prominent benchmark datasets, including MNIST, ETH 80, and Caltech. Our results demonstrate promising performance with minimal latency and power consumption compared to their electronic ONN counterparts. Additionally, we conducted several pertinent simulations, such as optical intensity to-latency conversion and synchronization. Of particular significance is the evaluation of the feature extraction layer, employing a Gabor filter bank, which stands to impact the practical deployment of diverse ONN architectures significantly.

TACNET: Temporal Audio Source Counting Network

In this paper, we introduce the Temporal Audio Source Counting Network (TaCNet), an innovative architecture that addresses limitations in audio source counting tasks. TaCNet operates directly on raw audio inputs, eliminating complex preprocessing steps and simplifying the workflow. Notably, it excels in real-time speaker counting, even with truncated input windows. Our extensive evaluation, conducted using the LibriCount dataset, underscores TaCNet's exceptional performance, positioning it as a state-of-the-art solution for audio source counting tasks. With an average accuracy of 74.18 percentage over 11 classes, TaCNet demonstrates its effectiveness across diverse scenarios, including applications involving Chinese and Persian languages. This cross-lingual adaptability highlights its versatility and potential impact.