Enhancing Vision-Language Models with Regional Understanding

Introducing RegionVLM to enhance vision-language models with regional understanding capabilities.