In this paper, we investigate implicit surface reconstruction methods based on deep learning, enhanced by multi-sensor data fusion, to improve the accuracy of 3D reconstruction in complex scenes. Existing single-sensor approaches often struggle with occlusions and incomplete observations. By fusing complementary information from multiple sensors (e.g., multiple cameras or a combination of cameras and depth sensors), our proposed framework alleviates the issue of missing or partial data and further increases reconstruction fidelity. We introduce a novel deep neural network that learns a continuous signed distance function (SDF) for scene geometry, conditioned on fused multi-sensor feature representations. The network seamlessly merges multi-modal data into a unified implicit representation, enabling precise and watertight surface reconstruction. We conduct extensive experiments on 3D datasets, demonstrating superior accuracy compared to single-sensor baselines and classical fusion methods. Quantitative and qualitative results reveal that multi-sensor fusion significantly improves reconstruction completeness and geometric detail, while our implicit approach provides smooth, high-resolution surfaces. Additionally, we analyze the influence of the number and diversity of sensors on reconstruction quality, the model’s ability to generalize to unseen data, and computational considerations. Our work highlights the potential of coupling deep implicit representations with multi-sensor fusion to achieve robust 3D reconstruction in challenging real-world conditions.

Achieving environmental sustainability has become an urgent priority in the era of rapid digital economic expansion, which presents both opportunities and challenges for environmental sustainable development. This study investigates the impact of digital economy (DIE) on environmental sustainability (ENS) through the dual dimensions of digital industrialization (DII) and industrial digitalization (IND), employing the wavelet-based quantile-on-quantile regression method to capture both quantile dependencies and temporal variations. The results reveal that DIE positively impacts ENS in the long term, while its short-term effects are mixed, with positive effects at lower and higher quantiles but negative impacts at mid-range quantiles of [0.35–0.45] and [0.65–0.7]. Specifically, DII exerts a predominantly negative short-term effect on ENS due to the environmental costs of digital infrastructure expansion, but turns positive in the long term as digital industrialization matures, especially in [0.85–0.95]. IND, conversely, exerts a consistently positive impact on ENS in both short- and long-term scenarios, highlighting its role in enhancing industrial efficiency and reducing emissions. These results emphasize the need for targeted policies, including prioritizing industrial digitalization, developing green infrastructure, and adopting phased digital development strategies to maximize environmental benefits.

This paper proposes a high-sensitivity U-shaped optical fiber sensor based on indium tin oxide (ITO) for surface plasmon resonance (SPR) sensing. Finite element simulations reveal that introducing ITO enhances the surface electric field strength by 1.15× compared to conventional designs, directly boosting sensitivity. The U-shaped structure optimizes evanescent wave–metal film interaction, further improving performance. In an external refractive index (RI) range of 1.334–1.374 RIU, the sensor achieves a sensitivity of 4333 nm/RIU (1.85× higher than traditional fiber sensors) and a figure of merit (FOM) of 21.7 RIU⁻¹ (1.68× improvement). Repeatability tests show a low relative standard deviation (RSD) of 0.4236% for RI measurements, with a maximum error of 0.00018 RIU, confirming excellent stability. The ITO coating’s strong adhesion ensures long-term reliability. With its simple structure, ease of fabrication, and superior sensitivity/FOM, this SPR sensor is well-suited for high-precision biochemical detection in intelligent sensing systems.

Large Language Models (LLMs) excel in diverse text generation tasks but still face limited controllability, opaque decision processes, and frequent hallucinations. This paper presents a structural causal intervention framework that models input–hidden–output dependencies through a structural causal model and performs targeted interventions on hidden representations. By combining counterfactual sample construction with contrastive training, our method enables precise control of style, sentiment, and factual consistency while providing explicit causal explanations for output changes. Experiments on three representative tasks demonstrate consistent and substantial improvements: style transfer accuracy reaches 92.3% (+7–14 percentage points over strong baselines), sentiment-controlled generation achieves 90.1% accuracy (+1.3–10.9 points), and multi-attribute conflict rates drop to 3.7% (a 40–60% relative reduction). Our method also improves causal attribution scores to 0.83–0.85 and human agreement rates to 87–88%, while reducing training and inference latency by 25–30% through sparse masking that modifies ≤10% of hidden units per attribute. These results confirm that integrating structural causal intervention with counterfactual training advances controllability, interpretability, and efficiency in LLM-based generation, offering a robust foundation for deployment in reliability-critical and resource-constrained applications.

Financial time series forecasting remains a persistent challenge due to the non-stationary nature, inherent noise, and multi-scale temporal dependencies present in market data. This paper presents MSTFNet, a multi-scale temporal fusion network that combines dilated causal convolutions with a frequency-enhanced sparse attention mechanism for improved financial prediction. The proposed architecture consists of three core components: a multi-scale dilated causal convolution module that extracts temporal patterns across different time horizons through parallel convolutional branches with varying dilation rates, a frequency-enhanced sparse attention mechanism that leverages Fast Fourier Transform to identify dominant periodic components and modulate attention weights accordingly, and an adaptive scale fusion gate that learns to dynamically combine representations from multiple temporal scales. Extensive experiments conducted on three public financial datasets (S&P 500, CSI 300, and NASDAQ Composite) spanning the period from January 2015 to December 2024 show two key results. First, consistent with near-efficient markets, the random-walk benchmark (${\widehat{y}}_{t+1}=y_t$) outperforms all the data-driven models on level-error metrics (MAE, RMSE, MAPE, and $R^2$), establishing the martingale as the binding lower bound on point-prediction error. Second, MSTFNet achieves the highest directional accuracy (DA) across all three indices—56.3% on the S&P 500 versus 50.0% for the martingale—representing a 6.3 percentage-point improvement that generates positive pre-cost returns in a trading strategy backtest. Among the eight data-driven baselines (LSTM, GRU, TCN, Transformer, Autoformer, FEDformer, PatchTST, and iTransformer), MSTFNet also achieves the lowest MAE, reducing it by 13.6% relative to the strongest data-driven baseline (iTransformer) on the S&P 500. These results confirm that integrating multi-scale temporal modeling with frequency-domain guidance extracts a real, if modest, directional signal from financial time series.

As the complexity of banking technology systems increases, the prevention of technological risk becomes an endless battle. Currently, most banks rely on the experience and subjective judgement of experts and employees to allocate resources for technological risk management, which does not effectively reduce the frequency of technology-related incidents. Through an analysis of mainstream risk management models, this study proposes a technology-based risk assessment system based on machine learning. It first identifies risk factors in bank IT, preprocesses the sample data, and uses different regression prediction models to train the processed data to build an intelligent assessment model. The experimental results indicated that the Genetic Algorithm–Backpropagation Neural Network model achieved the best performance. Based on assessment indicators, indicator weight values, and risk levels, commercial banks can develop targeted prevention and control measures by applying limited resources to the most critical corrective actions, thereby effectively reducing the frequency of technology-related incidents.