A Benchmark Study of Hybrid CNN-Transformer Architectures in Vision-Language Tasks. Emerging Science Research. 2025;3(01):36-49. Accessed April 23, 2026. https://emergingpub.com/index.php/sr/article/view/78