Steel surface defect classification is critical for industrial quality control, yet existing methods struggle to balance accuracy and efficiency for real-time deployment in vision-based sensor systems. This paper presents MobileSteelNet, a lightweight deep learning framework that introduces two novel modules: multi-scale feature fusion (MSFF), for integrating multi-stage features; and Cross-Interactive Efficient Multi-Scale Attention (CIEMA), which unifies inter-channel interaction, parallel multi-scale spatial extraction, and grouped efficient computation. Experiments on the NEU-DET dataset demonstrate that MobileSteelNet achieves 91.36% average accuracy, surpassing ResNet-50 (88.01%) and lightweight networks, including MobileNetV2 (86.08%). Notably, it achieves 93.70% accuracy on Scratch-type defects, representing an 82.12 percentage point improvement over baseline MobileNetV1. With a model size of only 8.2 MB, MobileSteelNet maintains superior performance while meeting lightweight deployment requirements, making it suitable for edge deployment in vision sensor systems for steel manufacturing.

Accurate matching between molecular structures and NMR spectra is an important task in automated structure elucidation. However, existing methods still face difficulties in jointly modeling multi-scale molecular topology and effectively exploiting the complementary information provided by paired ¹H and ¹³C NMR spectra. To address these limitations, we propose SpecMol-MatchNet, a multimodal matching framework that integrates a hybrid molecular graph encoder, branch-specific spectral feature learning, and residual multimodal fusion. In the molecular branch, attention-based graph interaction is combined with multi-scale neighborhood aggregation to capture structural cues at different receptive fields. In the spectral branch, branch-specific attention enhancement and joint gating are introduced to better exploit the complementary characteristics of paired ¹H and ¹³C spectra. The resulting molecular and spectral representations are integrated through a residual fusion module for final matching prediction. Experimental results on benchmark datasets demonstrate that SpecMol-MatchNet achieves consistently better overall performance than representative baseline methods.