Ancient Chinese architecture (ACA) has rich historical and cultural value, and the preservation of ACA can be enhanced by designing an efficient multi-label ACA image classification model. However, most existing methods mainly focus on convolutional neural networks to learn the local features of ancient architectural images, which cannot retain the positional information of each part of the ancient architecture and fail to leverage the semantic context effectively. This limitation leads to semantic confusion when dealing with similar ancient architectural structures. To solve these problems, we propose a novel coordinate-to-semantic attention network. It explicitly explores positional information and semantic context in an image to optimize the visual feature representation, thereby improving the accuracy of multi-label ACA image classification. Specifically, the coordinate attention mechanism is introduced to process information from different locations and capture long-range dependencies in the image. Subsequently, the semantic attention module realizes the interaction between visual and semantic information by establishing alignments between the two modalities, thereby producing semantically relevant visual representations for each label. Ultimately, the aggregated label features are processed to output the corresponding predictions. Experimental results on the 6-class ACA dataset and other datasets show that the proposed method significantly outperforms existing methods in terms of classification accuracy. Furthermore, the proposed method is of great significance in the protection and research application of ancient architectural heritage.
7 December 2024 in Journal of Real-Time Image Processing
In big data era, the special data with rare characteristics may be of great significations. However, it is very difficult to automatically search these samples from the massive and high-dimensional datasets and systematically evaluate them. The DoPS, our previous work [1], provided a search method of rare spectra with double-peaked profiles from massive and high-dimensional data of LAMOST survey. The identification of the results is mainly depended on visually inspection by astronomers. In this paper, as a follow-up study, a new lattice structure named SVM-Lattice is designed based on SVM(Support Vector Machine) and FCL(Formal Concept Lattice) and particularly applied in the recognition and evaluation of rare spectra with double-peaked profiles. First, each node in the SVM-Lattice structure contains two components: the intents are defined by the support vectors trained by the spectral samples with the specific characteristics, and the relevant extents are all the positive samples classified by the support vectors. The hyperplanes can be extracted from every lattice node and used as classifiers to search targets by categories. A generalization and specialization relationship is expressed between the layers, and higher layers indicate higher confidence of targets. Then, including a SVM-Lattice building algorithm, a pruning algorithm based on association rules, and an evaluation algorithm, the supporting algorithms are provided and analysed. Finally, for the recognition and evaluation of spectra with double-peaked profiles, several data sets from LAMOST survey are used as experimental dataset. The results exhibit good consistency with traditional methods, more detailed and accurate evaluations of classification results, and higher searching efficiency than other similar methods.