Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 2 appraisals
PloS one
An art style classification network integrating contrastive learning and counterfactual attention
Art style classification is a fundamental task for digital art analysis and intelligent cultural heritage management. However, existing methods often fail to deliver stable and interpretable results due to dispersed style cues, complex local textures, strong interference from background and subject content, and the difficulty of identifying regions that are causally relevant to style recognition. To address these issues, we propose a Multi Source Collaborative Style Network, termed MCS-Net, which jointly models global composition and local brushstroke textures within a unified framework. MCS-Net consists of four modules, namely a basic style feature encoding module, an attention generation module, a style contrastive learning module, and a counterfactual attention module. The attention generation module discovers multiple complementary discriminative regions. The contrastive learning module constructs a more discriminative embedding space to improve the separability of fine-grained styles. The counterfactual attention module builds counterfactual attention samples to explicitly estimate the true contribution of regions to predictions, thereby suppressing spurious activations, improving robustness, and providing counterfactual-supported evidence for model interpretation. Experiments on three public datasets, WikiArt, MultitaskPainting100k, and Pandora18k, demonstrate that MCS-Net outperforms representative baselines across standard evaluation metrics.
28 June 2026
Read appraisal →PloS one
Research on the semantic segmentation of Thangka images via an improved PIDNet.
In the digital preservation, restoration, and research of Thangka paintings, real-time semantic segmentation plays a crucial role in rapid image analysis. However, Thangka images exhibit intricate compositions, where principal figures often blend with backgrounds, ritual objects, and intricate ornaments, leading to blurred boundaries and fine details that challenge conventional segmentation methods in balancing accuracy and efficiency. To address this, we propose an improved PIDNet-based model incorporating ECA-Pag (Efficient Channel Attention-Path Aggregation) and LGFM (Local-Global Feature Fusion Module) modules, along with PConv3 (Partial Convolution 3), enhancing feature extraction and segmentation precision. Experimental results demonstrate that our model achieves 73.28% mIoU and mB-Fscore (mean Boundary F-score) of 40.01% on a custom Thangka dataset while maintaining 109.03 FPS (Frames Per Second), ensuring both high accuracy and real-time performance. Furthermore, evaluations on the Cityscapes benchmark confirm the model's generalization capability, outperforming baseline methods. This work provides an efficient and reliable solution for Thangka image segmentation, with potential applications in cultural heritage preservation and broader computer vision tasks.
17 May 2026
Read appraisal →