Conditional Prompt Learning for Vision-Language Models

Benchmark Model Rank Results
prompt-engineering-on-caltech-101CoCoOp#13Harmonic mean: 95.84
prompt-engineering-on-dtdCoCoOp#13Harmonic mean: 64.85
prompt-engineering-on-eurosatCoCoOp#13Harmonic mean: 71.21
prompt-engineering-on-fgvc-aircraftCoCoOp#14Harmonic mean: 27.74
prompt-engineering-on-food-101CoCoOp#12Harmonic mean: 90.99
prompt-engineering-on-imagenetCoCoOp#14Harmonic mean: 73.10
prompt-engineering-on-imagenet-aCoCoOp#7Top-1 accuracy %: 50.63
prompt-engineering-on-imagenet-rCoCoOP#8Top-1 accuracy %: 76.18
prompt-engineering-on-imagenet-sCoCoOp#8Top-1 accuracy %: 48.75
prompt-engineering-on-imagenet-v2CoCoOp#5Top-1 accuracy %: 64.07
prompt-engineering-on-oxford-102-flowerCoCoOp#13Harmonic mean: 81.71
prompt-engineering-on-oxford-iiit-pet-datasetCoCoOp#9Harmonic mean: 96.43
prompt-engineering-on-stanford-cars-1CoCoOp#13Harmonic mean: 72.01
prompt-engineering-on-sun397CoCoOp#13Harmonic mean: 78.27
prompt-engineering-on-ucf101CoCoOp#13Harmonic mean: 77.64