Ge, C., Qin, W. F. (2025). ScaleFormer architecture for scale invariant human pose estimation with enhanced mixed features. Scientific Reports. https://doi.org/10.1038/s41598-025-12620-4