Contrastive Feature Masking Open-Vocabulary Vision Transformer

Open paper
Benchmark Model Rank Results
open-vocabulary-object-detection-on-lvis-v1-0CFM-ViT–AP novel-LVIS base training: 33.9
open-vocabulary-object-detection-on-mscocoCFM-ViT–AP 0.5: 34.1

Rank counts only results with a code link.