Metasurfaces, capable of manipulating light at subwavelength scales, hold great potential for advancing optoelectronic applications. Generative models, particularly Generative Adversarial Networks (GANs), offer a promising approach for metasurface inverse design by efficiently navigating complex design spaces and capturing underlying data patterns. However, existing generative models struggle to achieve high electromagnetic fidelity and structural diversity. These challenges arise from the lack of explicit electromagnetic constraints during training, which hinders accurate structure‐to‐electromagnetic mapping, and the absence of mechanisms to handle one‐to‐many mappings dilemma, resulting in insufficient structural diversity. To address these issues, we propose the Anchor‐controlled Generative Adversarial Network (AcGAN), a novel framework that improves both electromagnetic fidelity and structural diversity. To achieve high electromagnetic fidelity, AcGAN proposes the Spectral Overlap Coefficient (SOC) for precise spectral fidelity assessment and develops AnchorNet, which provides real‐time physics‐guided feedback on electromagnetic performance to refine the structure‐to‐electromagnetic mapping. To enhance structural diversity, AcGAN incorporates a cluster‐guided controller that refines input processing and ensures multilevel spectral integration, guiding the generation process to explore multiple configurations. Empirical analysis shows that AcGAN reduces the Mean Squared Error (MSE) by 73 % compared to current state‐of‐the‐art and significantly expands the design space to generate diverse metasurface architectures that meet precise spectral demands.
The correlation between metasurface structures and their corresponding absorption spectra is inherently complex due to intricate physical interactions. Additionally, the reliance on Maxwell’s equations for simulating these relationships leads to extensive computational demands, significantly hindering rapid development in this area. Numerous researchers have employed artificial intelligence (AI) models to predict absorption spectra. However, these models often act as black boxes. Despite training high-performance models, it remains challenging to verify if they are fitting to rational patterns or merely guessing outcomes. To address these challenges, we introduce the Explainable Encoder–Prediction–Reconstruction (EEPR) framework, which separates the prediction process into feature extraction and spectra generation, facilitating a deeper understanding of the physical relationships between metasurface structures and spectra and unveiling the model’s operations at the feature level. Our model achieves a 66.23% reduction in average Mean Square Error (MSE), with an MSE of 2.843 compared to the average MSE of for mainstream networks. Additionally, our model operates approximately 500,000 times faster than traditional simulations based on Maxwell’s equations, with a time of seconds per sample, and demonstrates excellent generalization capabilities. By utilizing the EEPR framework, we achieve feature-level explainability and offer insights into the physical properties and their impact on metasurface structures, going beyond the pixel-level explanations provided by existing research. Additionally, we demonstrate the capability to adjust absorption by changing the metasurface at the feature level. These insights potentially empower designers to refine structures and enhance their trust in AI applications.
9 April 2024 in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
Realizing high-fidelity three-dimensional (3D) scene representation through holography presents a formidable challenge, primarily due to the unknown mechanism of the optimal hologram and huge computational load as well as memory usage. Herein, we propose a Physically Guided Generative Adversarial Network (PGGAN), which is the first generative model to transform the multi-view light field directly to holographic 3D content. PGGAN harmoniously fuses the fidelity of data-driven learning with the rigor of physical optics principles, ensuring a stable reconstruction quality across wide field of view, which is unreachable by current central-view-centric approaches. The proposed framework presents an innovative encoder-generator-discriminator, which is informed by a physical optics model. It benefits from the speed and adaptability of data-driven methods to facilitate rapid learning and effectively transfer to novel scenes, while its physics-based guidance ensures that the generated holograms adhere to holographic standards. A unique, differentiable physical model facilitates end-to-end training, which aligns the generative process with the “holographic space”, thereby improving the quality of the reconstructed light fields. Employing an adaptive loss strategy, PGGAN dynamically adjusts the influence of physical guidance in the initial training stages, later optimizing for reconstruction accuracy. Empirical evaluations reveal PGGAN’s exceptional ability to swiftly generate a detailed hologram in as little as 0.002 seconds, significantly eclipsing current state-of-the-art techniques in speed while maintaining superior angular reconstruction fidelity. These results demonstrate PGGAN’s effectiveness in producing high-quality holograms rapidly from multi-view datasets, advancing real-time holographic rendering significantly.
Research in artificial intelligence demonstrates the applicability and flexibility of the reinforcement learning (RL) technique for the dynamic job shop scheduling problem (DJSP). However, the RL-based method will always overfit to the training environment and cannot generalize well to novel unseen situations at deployment time, which is unacceptable in real-world production. For this reason, this paper proposes a highly generalizable reinforcement learning framework named Train Once For All (TOFA) for the dynamic job shop scheduling problem. The trivial and non-trivial states are distinguished when the DJSP is formulated as a semi-Markov decision process, defining the size-agnostic state, action, and reward function. A novel graph representation learning method based on attention mechanism and spatial pyramid pooling is implemented to compress the disjunctive graphs of differentsize DJSP into fixed-length feature vectors. Combining the proposed dynamic frame skipping and an improved prioritized experience replay method that considers the sample quality difference at different training phases. TOFA shows superb generalization capability, outperforms practically favored dispatching rules and even instance-by-instance training RL-based schedulers on various benchmark DJSP. Additionally, we proved that TOFA acquires a transferable scheduling policy that can be used to schedule a whole new DJSP without additional training.