Musical cultural heritage, as an important component of cultural heritage, possesses significant cultural value and inheritance significance. With the development of society and the passage of time, these precious traditional musical cultural heritages inevitably face the dilemma of gradual depletion or even disappearance. In the digital age, effectively protecting and inheriting these musical cultural heritages has become an urgent problem to be addressed. Therefore, this paper proposes an application method based on digital twin technology, exploring how to protect and inherit musical cultural heritages through digital twin technology. By leveraging digital twin technology, a virtual museum dedicated to showcasing the richness and historical connotations of music cultures is created, preserving and simulating the soundscapes of historical music eras. Through the integration of audio archives, 3D modeling, and interactive displays, users can immerse themselves in the experience of historical music in the digital space. This paper evaluates the feasibility and cultural preservation value of this digital music history museum through the creation of music digital twin technology instances and user survey feedback and discusses the prospects of digital twins in the field of musical cultural heritage.

This study focuses on the construction and evaluation of a high-quality Chinese Manchu music dataset designed to facilitate Artificial Intelligence (AI) research and applications within cultural heritage and ethnomusicology. Through a systematic collection and organization of diverse Manchu music resources, including folk songs, dance music, and ceremonial pieces, this dataset effectively represents the cultural breadth of Manchu music. The dataset includes digitized and preprocessed audio data, with comprehensive metadata annotations, such as essential information, musical features, and cultural context, creating a robust foundation for AI-based analysis. Experimental evaluations highlight the dataset’s utility across various AI-driven applications: in music classification, using a CNN model, an accuracy of 90% was achieved in the “folk ensemble” category, with an overall accuracy of 85.7% and a precision of 82.3%. For music generation, a Generative Adversarial Network (GAN) model yielded a quality score of 7.8/10 and a Fréchet Audio Distance (FAD) of 0.32. In emotion recognition, the Random Forest model achieved 87% accuracy in identifying the emotion “joy”. These results underscore the dataset’s potential in supporting digital preservation and expanding AI applications in ethnic music classification, generation, and emotional analysis, contributing to both cultural heritage preservation and AI advancement in ethnomusicology.

This study focuses on advancing the field of remote sensing image target detection, addressing challenges such as small target detection, complex background handling, and dense target distribution. We propose solutions based on enhancing the YOLOv7 algorithm. Firstly, we improve the multi-scale feature enhancement (MFE) method of YOLOv7, enhancing its adaptability and precision in detecting small targets and complex backgrounds. Secondly, we design a modified YOLOv7 global information DP-MLP module to effectively capture and integrate global information, thereby improving target detection accuracy and robustness, especially in handling large-scale variations and complex scenes. Lastly, we explore a semi-supervised learning model (SSLM) target detection algorithm incorporating unlabeled data, leveraging information from unlabeled data to enhance the model’s generalization ability and performance. Experimental results demonstrate that despite the outstanding performance of YOLOv7, the mean average precision (MAP) can still be improved by 1.9%. Specifically, under testing on the TGRS-HRRSD-Dataset, the MFE and DP-MLP models achieve MAP values of 93.4% and 93.1%, respectively. Across the NWPU VHR-10 dataset, the three models achieve MAP values of 93.1%, 92.1%, and 92.2%, respectively. Significant improvements are observed across various metrics compared to the original model. This study enhances the adaptability, accuracy, and generalization of remote sensing image object detection.

Music cultural heritage, as an important component of cultural heritage, possesses significant cultural value and inheritance significance. With the development of society and the passage of time, these precious traditional music cultural heritages inevitably face the dilemma of gradual depletion or even disappearance. In the digital age, effectively protecting and inheriting these music cultural heritages has become an urgent problem to be addressed. Therefore, this paper proposes an application method based on digital twin technology, exploring how to protect and inherit music cultural heritages through digital twin technology. By leveraging digital twin technology, a virtual museum dedicated to showcasing the richness and historical connotations of music cultures is created, preserving and simulating the soundscapes of historical music eras. Through the integration of audio archives, 3D modeling, and interactive displays, users can immerse themselves in the experience of historical music in the digital space. This paper evaluates the feasibility and cultural preservation value of this digital music history museum through the creation of music digital twin technology instances and user survey feedback, and discusses the prospects of digital twins in the field of music cultural heritage.

Medical image segmentation is a challenging task due to limited annotated data, complex lesion boundaries, and the inherent variability in medical images. These challenges make accurate and robust segmentation crucial for clinical applications. In this study, we propose the Uncertainty-Driven Auxiliary Mean Teacher (UDAMT) model, a novel semi-supervised framework specifically designed for skin lesion segmentation. Our approach employs a dual-headed teacher–student architecture with an uncertainty-guided mechanism, enhancing feature learning and boundary precision. Extensive experiments on the ISIC 2016, ISIC 2017, and ISIC 2018 datasets demonstrate that UDAMT achieves significant improvements over state-of-the-art methods, with increases of 1.17 percentage points in the Dice coefficient and 1.31 percentage points in mean Intersection over Union (mIoU) under low-label settings (5% labeled data). Furthermore, UDAMT requires 12.9 M parameters, which is slightly higher than the baseline model (12.5 M) but significantly lower than MT (14.8 M) and UAMT (15.2 M). It also achieves an inference time of 25.7 ms per image, ensuring computational efficiency. Ablation studies validate the contributions of each component, and cross-dataset evaluations on the PH2 benchmark confirm robustness to small lesions. This work provides a scalable and efficient solution for semi-supervised medical image segmentation, balancing accuracy, efficiency, and clinical applicability.

한국멀티미디어학회(The Journal of Curriculum Evaluation)