# Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA Tasks? A: Self-Train on Unlabeled Images!

> Research article (2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023) · cited 11× · AI/ML

**Wikidata**: [openalex:W4386065359](https://www.wikidata.org/wiki/openalex:W4386065359)  
**Source**: https://4ort.xyz/entity/q-how-to-specialize-large-vision-language-models-to-data-scarce-vqa-tasks-a-self-train-on-unlabeled-images