Search Results (96)

Diffuse Optical Tomography (DOT) is a non-invasive medical imaging technique that makes use of Near-Infrared (NIR) light to recover the spatial distribution of optical coefficients in biological tissues for diagnostic purposes. Due to the intense scattering of light within tissues, the reconstruction process inherent to DOT is severely ill-posed. In this paper, we propose to tackle the ill-conditioning by learning a prior over the solution space using an autoencoder-type neural network. Specifically, the decoder part of the autoencoder is used as a generative model. It maps a latent code to estimated physical parameters given in input to the forward model. The latent code is itself the result of an optimization loop which minimizes the discrepancy of the solution computed by the forward model with available observations. The structure and interpretability of the latent space are enhanced by minimizing the rank of its covariance matrix, thereby promoting more effective utilization of its information-carrying capacity. The deep learning-based prior significantly enhances reconstruction capabilities in this challenging domain, demonstrating the potential of integrating advanced neural network techniques into DOT. Full article

The integration of high-penetration renewable energy sources (RESs) into transmission networks introduces profound uncertainty that challenges traditional infrastructure planning approaches. Existing transmission expansion planning (TEP) models either rely on static scenario sets or over-conservative worst-case assumptions, failing to capture the operational stress triggered by rare but structurally impactful renewable behaviors. This paper proposes a novel bi-level optimization framework for transmission planning under adversarial uncertainty, coupling a distributionally robust upper-level investment model with a lower-level operational response embedded with physics and market constraints. The uncertainty space was not exogenously fixed, but instead dynamically generated through a physics-informed spatiotemporal generative adversarial network (PI-ST-GAN), which synthesizes high-risk renewable and load scenarios designed to maximally challenge the system’s resilience. The generator was co-trained using a composite stress index—combining expected energy not served, loss-of-load probability, and marginal congestion cost—ensuring that each scenario reflects both physical plausibility and operational extremity. The resulting bi-level model was reformulated using strong duality, and it was decomposed into a tractable mixed-integer structure with embedded adversarial learning loops. The proposed framework was validated on a modified IEEE 118-bus system with high wind and solar penetration. Results demonstrate that the GAN-enhanced planner consistently outperforms deterministic and stochastic baselines, reducing renewable curtailment by up to 48.7% and load shedding by 62.4% under worst-case realization. Moreover, the stress investment frontier exhibits clear convexity, enabling planners to identify cost-efficient resilience strategies. Spatial congestion maps and scenario risk-density plots further illustrate the ability of adversarial learning to reveal latent structural bottlenecks not captured by conventional methods. This work offers a new methodological paradigm, in which optimization and generative AI co-evolve to produce robust, data-aware, and stress-responsive transmission infrastructure designs. Full article

Generative image models have revolutionized artificial intelligence by enabling the synthesis of high-quality, realistic images. These models utilize deep learning techniques to learn complex data distributions and generate novel images that closely resemble the training dataset. Recent advancements, particularly in diffusion models, have led to remarkable improvements in image fidelity, diversity, and controllability. In this work, we investigate the application of a conditional latent diffusion model in the healthcare domain. Specifically, we trained a latent diffusion model using unlabeled histopathology images. Initially, these images were embedded into a lower-dimensional latent space using a Vector Quantized Generative Adversarial Network (VQ-GAN). Subsequently, a diffusion process was applied within this latent space, and clustering was performed on the resulting latent features. The clustering results were then used as a conditioning mechanism for the diffusion model, enabling conditional image generation. Finally, we determined the optimal number of clusters using cluster validation metrics and assessed the quality of the synthetic images through quantitative methods. To enhance the interpretability of the synthetic image generation process, expert input was incorporated into the cluster assignments. Full article

Wake effects significantly reduce efficiency and increase structural loads in wind farms. Therefore, accurate and computationally efficient models are crucial for wind farm layout optimization and operational control. High-fidelity computational fluid dynamics (CFD) simulations, while accurate, are too slow for these tasks, whereas faster analytical models often lack dynamic fidelity and 3D detail, particularly under complex conditions. Existing data-driven surrogate models based on neural networks often struggle with the high dimensionality of the flow field and scalability to large wind farms. This paper proposes a novel data-driven surrogate modeling framework to bridge this gap, leveraging Neural Networks (NNs) trained on data from the high-fidelity SOWFA (simulator for wind farm applications) tool. A physics-inspired NN architecture featuring an autoencoder for spatial feature extraction and latent space dynamics for temporal evolution is introduced, motivated by the time–space decoupling structure observed in the Navier–Stokes equations. To address scalability for large wind farms, a “wind box” decomposition strategy is employed. This involves training separate NN models on smaller, canonical domains (with and without turbines) that can be stitched together to represent larger farm layouts, significantly reducing training data requirements compared to monolithic farm simulations. The development of a batch simulation interface for SOWFA to generate the required training data efficiently is detailed. Results demonstrate that the proposed surrogate model accurately predicts the 3D dynamic wake evolution for single-turbine and multi-turbine configurations. Specifically, average velocity errors (quantified as RMSE) are typically below 0.2 m/s (relative error < 2–5%) compared to SOWFA, while achieving computational accelerations of several orders of magnitude (simulation times reduced from hours to seconds). This work presents a promising pathway towards enabling advanced, model-based optimization and control of large wind farms. Full article

The accelerated penetration of digital technology into natural ecosystems has led to the digital transformation of forest ecological spaces. Smart forestry, as a key pathway for digital-intelligence-enabled ecological governance, plays an important role in global sustainable development and multi-level governance. However, due to differences in functional positioning, resource capacity, and policy translation mechanisms, semantic shifts and disconnections arise between central policies, local policies, and practical implementation, thereby affecting policy execution and governance effectiveness. Fujian Province has been identified as a key pilot region for smart forestry practices in China, owing to its early adoption of informatization strategies and distinctive ecological conditions. This study employed the Latent Dirichlet Allocation (LDA) topic modeling method to construct a corpus of smart forestry texts, including central policies, local policies, and local media reports from 2010 to 2025. Seven potential themes were identified and categorized into three overarching dimensions: technological empowerment, governance mechanisms, and ecological goals. The results show that central policies emphasize macro strategy and ecological security, local policies focus on platform construction and governance coordination, and local practice features digital innovation and ecological value transformation. Three transmission paths are summarized to support smart forestry policy optimization and inform digital ecological governance globally. Full article

The use of recurrent neural networks has proven effective in time series prediction tasks such as weather. However, their use in resource-limited systems such as MCUs presents difficulties in terms of both size and stability with longer prediction windows. In this context, we propose a variant of the LSTM model, which we call SE-LSTM (Single Embedding LSTM), which uses embedding techniques to vectorially represent seasonality and latent patterns through variables such as temperature and humidity. The proposal is systematically compared in two parts: The first compares it against other reference architectures such as CNN-LSTM, TCN, LMU, and TPA-LSTM. The second stage, which includes implementation, compares it against the CNN-LSTM, LSTM, and TCN networks. Metrics such as the MAE and MSE are used along with the network weight, a crucial aspect for MCUs such as an ESP32 or Raspberry Pi Pico. An analysis of the memory usage, energy consumption, and generalization across different regions is also included. The results show that the use of embedding optimizes the network space without sacrificing the performance, which is crucial for edge computing. This research is part of a larger project, which focuses on improving agricultural monitoring systems. Full article

We propose Phase-Adaptive Federated Learning (PAFL), a novel framework for privacy-preserving personalized travel itinerary generation that dynamically balances privacy and utility through a phase-dependent aggregation mechanism inspired by phase-change materials. (1) PAFL’s primary objective is to dynamically optimize the privacy–utility trade-off in federated travel recommendation systems through phase-adaptive anonymization. The phase parameter φ ∈ [0, 1] operates as a tunable control variable that continuously adjusts the latent space geometry between differentially private (φ→1) and utility-optimized (φ→0) representations via a thermodynamic-inspired transformation. Conventional federated learning approaches often rely on static privacy-preserving techniques, which either degrade recommendation quality or inadequately protect sensitive user data; PAFL addresses this limitation through three key innovations: a latent-space phase transformer, a differential privacy-gradient inverter with mathematically provable reconstruction bounds (εt ≤ 1.0), and a lightweight sequential transformer. (2) PAFL’s core innovation lies in its phase-adaptive mechanism that dynamically balances privacy preservation through differential privacy and utility maintenance via gradient inversion, governed by the tunable phase parameter φ. Experimental results demonstrate statistically significant improvements, with 18.7% higher HR@10 (p < 0.01) and 62% lower membership inference risk compared to state-of-the-art methods, while maintaining εtotal < 2.3 over 100 training rounds. The framework advances federated learning for sensitive recommendation tasks by establishing a new paradigm for adaptive privacy–utility optimization. Full article

Molecular optimization plays a pivotal role in many domains since it holds promise for improving the properties of lead molecules. The advent of artificial intelligence (AI)-driven molecular optimization has revolutionized lead optimization workflows, which have significantly accelerated the development of drug candidates. However, AI models are also confronted with new challenges in practical molecular optimization, such as high-dimensional chemical space and data sparsity issues. This paper initially highlights the inherent benefits of molecular optimization in terms of optimizing the properties and maintaining the structural similarity of lead molecules, thereby highlighting its critical role in drug discovery. The next section systematically categorizes and analyzes existing AI-aided molecular optimization methods, comprising iterative search in discrete chemical space, end-to-end generation in continuous latent space, and iterative search in continuous latent space methods. Finally, we discuss the key challenges in AI-aided molecular optimization methods, including molecular representations, dataset selection, the properties to be optimized, and optimization algorithms, while proposing potential solutions and future research directions. In summary, this review provides a comprehensive analysis of existing representative AI-aided molecular optimization methods, thereby offering guidance for future research directions. Full article

Diffusion models have made remarkable progress in image generation, leading to advancements in the field of image editing. However, balancing editability with faithfulness remains a significant challenge. Motivated by the fact that more novel content will be generated when larger variance noise is applied to the image, in this paper, we propose an Enhanced Non-isotropic Gaussian Diffusion Model (ENGDM) for progressive image editing, which introduces independent Gaussian noise with varying variances to each pixel based on its editing needs. To enable efficient inference without retraining, ENGDM is rectified into an isotropic Gaussian diffusion model (IGDM) by assigning different total diffusion times to different pixels. Furthermore, we introduce reinforced text embeddings, using a novel editing reinforcement loss in the latent space to optimize text embeddings for enhanced editability. And we introduce optimized noise variances by employing a structural consistency loss to dynamically adjust the denoising time steps for each pixel for better faithfulness. Experimental results on multiple datasets demonstrate that ENGDM achieves state-of-the-art performance in image-editing tasks, effectively balancing faithfulness to the source image and alignment with the desired editing target. Full article

Hybrid evolutionary approaches have gained significant attention for solving complex optimization problems, but their potential for optimizing the low-dimensional latent space of generative adversarial networks (GANs) remains underexplored. This paper proposes a novel improved crisscross optimization (ICSO) algorithm, a hybrid evolutionary approach that integrates crisscross optimization and perturbation mechanisms to find the suitable latent vector. The ICSO algorithm treats the quality and diversity as separate objectives, balancing them through a normalization strategy, while a gradient regularization term (i.e., GP) is introduced into the discriminator’s objective function to stabilize training and mitigate gradient-related issues. By combining the global and local search capabilities of particle swarm optimization (PSO) with the rapid convergence of crisscross optimization, ICSO efficiently explores and exploits the latent space. The extensive experiments demonstrate that ICSO outperforms state-of-the-art algorithms in optimizing the latent space of various classical GANs across multiple datasets. Furthermore, the practical applicability of ICSO is validated through its integration with StyleGAN3 for generating unmanned aerial vehicle (UAV) images, showcasing its effectiveness in real-world engineering applications. This work not only advances the field of GAN optimization but also provides a robust framework for applying hybrid evolutionary algorithms to complex generative modeling tasks. Full article

Multi-view clustering demonstrates strong performance in various real-world applications. However, real-world data often contain incomplete and uncoupled views. Missing views can lead to the loss of latent information, and uncoupled views create obstacles for cross-view learning. Existing methods rarely consider incomplete and uncoupled multi-view data simultaneously. To address these problems, a novel method called Tensor-based Uncoupled and Incomplete Multi-view Clustering (TUIMC) is proposed to effectively handle incomplete and uncoupled data. Specifically, the proposed method recovers missing samples in a low-dimensional feature space. Subsequently, the self-representation matrices are paired with the optimal views through permutation matrices. The coupled self-representation matrices are integrated into a third-order tensor to explore high-order information of multi-view data. An efficient algorithm is designed to solve the proposed model. Experimental results on five widely used benchmark datasets show that the proposed method exhibits superior clustering performance on incomplete and uncoupled multi-view data. Full article

With the accelerated transition of the global energy structure towards decarbonization, the share of PV power generation in the power system continues to rise. IEA predicts PV will account for 80% of new global renewable installations during 2025–2030. However, latent faults emerging from the long-term operation of photovoltaic (PV) power plants significantly compromise their operational efficiency. The existing EL detection methods in PV plants face challenges including grain boundary interference, probe band artifacts, non-uniform luminescence, and complex backgrounds, which elevate the risk of missing small defects. In this paper, we propose a high-precision defect detection method based on BiFDRep-YOLOv8n for small target defects in photovoltaic (PV) power plants, aiming to improve the detection accuracy and real-time performance and to provide an efficient solution for the intelligent detection of PV power plants. Firstly, the visual transformer RepViT is constructed as the backbone network, based on the dual-path mechanism of Token Mixer and Channel Mixer, to achieve local feature extraction and global information modeling, and combined with the structural reparameterization technique, to enhance the sensitivity of detecting small defects. Secondly, for the multi-scale characteristics of defects, the neck network is optimized by introducing a bidirectional weighted feature pyramid network (BiFPN), which adopts an adaptive weight allocation strategy to enhance feature fusion and improve the characterization of defects at different scales. Finally, the detection head part uses DyHead-DCNv3, which combines the triple attention mechanism of scale, space, and task awareness, and introduces deformable convolution (DCNv3) to improve the modeling capability and detection accuracy of irregular defects. Full article

Variational autoencoders (VAEs) have emerged as a promising tool for modeling volatility surfaces, with particular significance for generating synthetic implied volatility scenarios that enhance risk management capabilities. This study evaluates VAE performance using synthetic volatility surfaces, chosen specifically for their arbitrage-free properties and clean data characteristics. Through a comprehensive comparison with traditional methods including thin-plate spline interpolation, parametric models (SABR and SVI), and deterministic autoencoders, we demonstrate that our VAE approach with latent space optimization consistently outperforms existing methods, particularly in scenarios with extreme data sparsity. Our findings show that accurate, arbitrage-free surface reconstruction is achievable using only 5% of the original data points, with errors 7–12 times lower than competing approaches in high-sparsity scenarios. We rigorously validate the preservation of critical no-arbitrage conditions through probability distribution analysis and total variance strip non-intersection tests. The framework we develop overcomes traditional barriers of limited market data by generating over 13,500 synthetic surfaces for training, compared to typical market availability of fewer than 100. These capabilities have important implications for market risk analysis, derivatives pricing, and the development of more robust risk management frameworks, particularly in emerging markets or for newly introduced derivatives where historical data are scarce. Our integration of machine learning with financial theory constraints represents a significant advancement in volatility surface modeling that balances statistical accuracy with financial relevance. Full article

Recent studies have demonstrated that machine learning-based generative models can create novel molecules with desirable properties. Among them, Conditional Variational Autoencoder (CVAE) is a powerful approach to generate molecules with desired physiochemical and pharmacological properties. However, the CVAE’s latent space is still a black-box, making it difficult to understand the relationship between the latent space and molecular properties. To address this issue, we propose the Interpretable Conditional Variational Autoencoder (ICVAE), which introduces a modified loss function that correlates the latent value with molecular properties. ICVAE established a linear mapping between latent variables and molecular properties. This linearity is not only crucial for improving interpretability, by assigning clear semantic meaning to latent dimensions, but also provides a practical advantage. It enables direct manipulation of molecular attributes through simple coordinate shifts in latent space, rather than relying on opaque, black-box optimization algorithms. Our experimental results show that the ICVAE can linearly relate one or multiple molecular properties with the latent value and generate molecules with precise properties by controlling the latent values. The ICVAE’s interpretability allows us to gain insight into the molecular generation process, making it a promising approach in drug discovery and material design. Full article

Travel mode identification (TMI) plays a crucial role in intelligent transportation systems by accurately identifying travel modes from Global Positioning System (GPS) trajectory data. Given that trajectory data inherently exhibit spatial and kinematic patterns that complement each other, recent TMI methods generally combine these characteristics through image-based projections or direct concatenation. However, such approaches achieve only shallow fusion of these two types of features and cannot effectively align them into a shared latent space. To overcome this limitation, we introduce multi-view contrastive fusion (MVCF)-TMI, a novel TMI framework that enhances identification accuracy and model generalizability by aligning spatial and kinematic views through multi-view contrastive learning. Our framework employs multi-view learning to separately extract spatial and kinematic features, followed by an inter-view contrastive loss to optimize feature alignment in a shared subspace. This approach enables cross-view semantic understanding and better captures complementary information across different trajectory representations. Extensive experiments show that MVCF-TMI outperforms baseline methods, achieving 86.45% accuracy on the GeoLife dataset. The model also demonstrates strong generalization by transferring knowledge from pretraining on the large-scale GeoLife dataset to the smaller SHL dataset. Full article