Due to low contrast, chromatic aberration, and generally small objects in underwater environments, a new underwater fish detection model, BGLE-YOLO, is proposed to investigate automated methods dedicated to accurately detecting underwater objects in images. The model has small parameters and low computational effort and is suitable for edge devices. First, an efficient multi-scale convolutional EMC module is introduced to enhance the backbone network and capture the dynamic changes in targets in the underwater environment. Secondly, a global and local feature fusion module for small targets (BIG) is integrated into the neck network to preserve more feature information, reduce error information in higher-level features, and increase the model’s effectiveness in detecting small targets. Finally, to prevent the detection accuracy impact due to excessive lightweighting, the lightweight shared head (LSH) is constructed. The reparameterization technique further improves detection accuracy without additional parameters and computational cost. Experimental results of BGLE-YOLO on the underwater datasets DUO (Detection Underwater Objects) and RUOD (Real-World Underwater Object Detection) show that the model achieves the same accuracy as the benchmark model with an ultra-low computational cost of 6.2 GFLOPs and an ultra-low model parameter of 1.6 MB.

Pt is helpful for preventing Pd aggregation and promoting stability. Pd–3Pt/γ-Al ₂ O ₃ shows the best catalytic activity for CH ₄ activation in the presence of water.

Wheelset bearing fault diagnosis becomes challenging under time-varying train speeds, where traditional diagnostic methods often fail to remain effective. To address this issue, a multi-source physical feature graph attention network based on vibration–rotational speed data fusion is proposed to identify wheelset bearing failures under time-varying speed conditions. The method fuses three types of information to enhance rotational speed adaptability. First, a two-channel convolutional neural network is constructed to extract features from vibration signals and key-phase signals. Next, a self-attention mechanism combined with global average pooling is introduced to fuse the two feature sources. To further obtain effective information, a recursive graph auto-encoder exploits the recursiveness of vibration to generate inter-sample relationship graphs, enhancing feature interpretability. Finally, a graph attention network is used to fuse the extracted features with the recurrence graph. Experiments demonstrate that the proposed method achieves higher accuracy and rotational speed adaptability under time-varying speed conditions.