Chen, X., Mao, D., Ke, J. (2025). Real-world super-resolution with VLM-based degradation prior learning. Scientific Reports. https://doi.org/10.1038/s41598-025-14581-0