In recent years, accelerated global climate change has precipitated an increased frequency of wildfire events, with their devastating impacts on ecological systems and human populations becoming increasingly significant. Satellite remote sensing technology, leveraging its extensive spatial coverage and real-time monitoring capabilities, has emerged as a pivotal approach for wildfire early warning and comprehensive disaster assessment. To effectively detect subtle smoke signatures while minimizing background interference in remote sensing imagery, this paper introduces a novel dual-branch attention framework (CSFAttention) that synergistically integrates channel–spatial refinement with frequency spectral analysis to aggregate smoke features in remote sensing images. The channel–spatial branch implements an innovative triple-pooling strategy (incorporating average, maximum, and standard deviation pooling) across both channel and spatial dimensions to generate complementary descriptors that enhance distinct statistical properties of smoke representations. Concurrently, the frequency branch explicitly enhances high-frequency edge patterns, which are critical for distinguishing subtle textural variations characteristic of smoke plumes. The outputs from these complementary branches are fused through element-wise summation, yielding a refined feature representation that optimizes channel dependencies, spatial saliency, and spectral discriminability. The CSFAttention module is strategically integrated into the bottleneck structures of the ResNet architecture, forming a specialized deep network specifically designed for robust smoke recognition. Experimental validation on the USTC_SmokeRS dataset demonstrates that the proposed CSFResNet achieves recognition accuracy of 96.84%, surpassing existing deep networks for RS smoke recognition.
Link prediction in bipartite networks is a challenging task due to their distinct structural characteristics, where edges only exist between nodes of different types. Most existing methods are based on structural similarity, assigning similarity scores to node pairs under the assumption that a higher similarity corresponds to a higher likelihood of connection. Local structural methods, in particular, are widely favored for their simplicity, interpretability, and computational efficiency. However, real-world bipartite networks often exhibit highly heterogeneous node degree distributions, which introduce biases and undermine the effectiveness of traditional local structure-based methods. To address this issue, we propose a novel link prediction framework that explicitly adjusts for the degree heterogeneity of intermediate nodes between unconnected node pairs and incorporates their influence within local connection patterns formed around these pairs. Furthermore, our framework differentiates between the roles of same-type and cross-type nodes by leveraging quadrangle graphs between unconnected nodes. This approach allows for a more nuanced capture of unique properties of bipartite networks and effectively mitigates the inherent degree bias commonly observed in such networks, resulting in considerable improvements in prediction accuracy. Experimental results on ten diverse bipartite networks demonstrate that our framework achieves competitive and robust performance compared to nineteen state-of-the-art link prediction methods.