FOCUS: Unified Vision-Language Modeling for Interactive Editing Driven by Referential Segmentation

Benchmark Model Rank Results
referring-expression-segmentation-on-refcoco-plus-testaFOCUS–Overall IoU: 84.1
referring-expression-segmentation-on-refcoco-plus-valFOCUS–Overall IoU: 78.5
referring-expression-segmentation-on-refcoco-valFOCUS–Overall IoU: 84.1
visual-question-answering-on-mm-vetFOCUS–GPT-4 score: 50.2