This study proposes an enhanced method for natural earthquake and artificial explosion recognition, which comprises two parts, namely the multiscale fuzzy entropy (MFE) feature extraction of complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and the non-dominated sorting genetic algorithm III (NSGAIII) optimization of the one-dimensional convolutional neural network (1D-CNN). CEEMDAN decomposes earthquake signals into initial functions (intrinsic mode functions) and extracts fuzzy entropy features to construct a discriminative time-frequency representation. The hyperparameters of 1D-CNN (minimum batch size, initial learning rate, and learning rate drop factor) were optimized by NSGAIII, using a dual objective function to minimize mean squared error and maximize R2. Tests on 1000 earthquake events (883 earthquakes and 117 explosions) showed that the model has an accuracy of 97.82%, which is better than traditional networks (1D-CNN, generalized regression neural network, probabilistic neural network, back propagation neural network, and radial basis function neural network) and has better regression indicators (mean absolute error = 0.0795, root mean squared error = 0.1302, R2 = 0.7361). The Adam optimization algorithm achieved peak performance (99.50%), significantly surpassing SGD-M and RMSprop. This framework effectively solves the small sample and high-dimensional classification problems in earthquake monitoring and improves the automatic event detection capability of the early warning system.

Video deblurring faces a fundamental challenge, as blur degradation comprehensively affects frames by not only causing detail loss but also severely distorting structural information. This dual degradation across low- and high-frequency domains makes it challenging for existing methods to simultaneously restore both structural and detailed information through a unified approach. To address this issue, we propose a wavelet-based, blur-aware decoupled network (WBDNet) that innovatively decouples structure reconstruction from detail enhancement. Our method decomposes features into multiple frequency bands and employs specialized restoration strategies for different frequency domains. In the low-frequency domain, we construct a multi-scale feature pyramid with optical flow alignment. This enables accurate structure reconstruction through bottom-up progressive feature fusion. For high-frequency components, we combine deformable convolution with a blur-aware attention mechanism. This allows us to precisely extract and merge sharp details from multiple frames. Extensive experiments on benchmark datasets demonstrate the superior performance of our method, particularly in preserving structural integrity and detail fidelity.