# High-Precision Visual Question Answering

VideoRefer-7B is a multimodal model based on a video large language model, focusing on spatio-temporal object understanding tasks.

Llama-3.2V-11B-cot is a visual-language model capable of spontaneous and systematic reasoning, developed based on the LLaVA-CoT framework.

xGen-MM is a series of the latest foundational large multimodal models developed by Salesforce AI Research. It is improved based on the successful design of the BLIP series, providing more powerful multimodal processing capabilities.

InternLM-XComposer2 is a vision-language large model developed based on InternLM2, featuring outstanding image-text understanding and creation capabilities.