EVF-SAM: Early Vision-Language Fusion for Text-Prompted Segment Anything Model

Benchmark Model Rank Results
referring-expression-segmentation-on-refcocoEVF-SAM#6Overall IoU: 82.4
referring-expression-segmentation-on-refcoco-3EVF-SAM#4Overall IoU: 76.5
referring-expression-segmentation-on-refcoco-4EVF-SAM#4Overall IoU: 80
referring-expression-segmentation-on-refcoco-5EVF-SAM#3Overall IoU: 71.9
referring-expression-segmentation-on-refcoco-8EVF-SAM#4Overall IoU: 84.2
referring-expression-segmentation-on-refcoco-9EVF-SAM#4Overall IoU: 80.2
referring-expression-segmentation-on-refcocogEVF-SAM#5Overall IoU: 78.2
referring-expression-segmentation-on-refcocog-1EVF-SAM#5Overall IoU: 78.3