Music genre classification (MGC) is the basis for the efficient organization, retrieval, and recommendation of music resources, so it has important research value. Convolutional neural networks (CNNs) have been widely used in MGC and achieved excellent results. However, CNNs cannot model global features well due to the influence of the local receptive field; these global features are crucial for classifying music signals with temporal properties. Transformers can capture long-range dependencies within an image thanks to adopting the self-attention mechanism. Nevertheless, there are still performance and computational cost gaps between Transformers and existing CNNs. In this paper, we propose a hybrid architecture (CNN-TE) based on CNN and Transformer encoder for MGC. Specifically, we convert the audio signals into mel spectrograms and feed them into a hybrid model for training. Our model employs a CNN to initially capture low-level and localized features from the spectrogram. Subsequently, these features are processed by a Transformer encoder, which models them globally to extract high-level and abstract semantic information. This refined information is then classified using a multi-layer perceptron. Our experiments demonstrate that this approach surpasses many existing CNN architectures when tested on the GTZAN and FMA datasets. Notably, it achieves these results with fewer parameters and a faster inference speed.

This article introduces a method for detecting ice floes in the Yellow River based on unmanned aerial vehicle (UAV) remote sensing images and deep learning technology. First, the article establishes a semantic segmentation dataset for Yellow River ice floes, filling the current data gap in this field. It designs a data augmentation scheme to improve the algorithm’s overfitting phenomenon. By analyzing the features of ice jam UAV remote sensing images and experimental verification, this article employs the U-Net algorithm as the foundation. By introducing the ConvNeXt-U feature extraction network, Squeeze-and-Excitation Network attention mechanism, and designing a hybrid loss function, the algorithm’s feature extraction ability is enhanced, improving the detection effect of small targets and overall improving the detection metrics of various semantic segmentation algorithms. Experimental results show that the proposed ConvNeXt-SE-U-Net increases accuracy by 16.05%, precision by 38.89%, and mean intersection over union by 30.35%. This method can accomplish high-precision ice floe semantic segmentation tasks in UAV remote sensing images, which is of great significance for Yellow River management.