Preprint
Aug 2026
EndoVLM: An Endoscopy Vision-Language Pre-training Model via Anatomy-Guided Sparsity and Progressive Alignment
EndoVLM is presented, a novel vision-language FM pre-trained on over 348K endoscopic examinations, each pairing a clinical report with its corresponding image collection, which outperforms existing foundation models and remains competitive with task-specific methods.
Zhen-Yu Yi, Jian-Wei Xu, Yue Hu et al.
· 0 citations