Unified Object Referring and Segmentation for Pixel-Level Visual Reasoning

UniPixel is a unified MLLM for pixel-level vision-language understanding. It flexibly supports a variety of fine-grained tasks, including image/video segmentation, regional understanding, and a novel PixelQA task that jointly requires object-centric referring, segmentation, and question-answering in videos. Please open an issue if you meet any problems.