Learning Hierarchical Prompt with Structured Linguistic Knowledge for Vision-Language Models

Benchmark Model Rank Results
prompt-engineering-on-oxford-102-flowerHPT#4Harmonic mean: 87.16
prompt-engineering-on-imagenet-v2HPT#5Top-1 accuracy %: 65.25
prompt-engineering-on-ucf101HPT#7Harmonic mean: 83.16
prompt-engineering-on-caltech-101HPT#10Harmonic mean: 96.65
prompt-engineering-on-imagenet-aHPT#10Top-1 accuracy %: 50.85
prompt-engineering-on-sun397HPT#10Harmonic mean: 80.88
prompt-engineering-on-fgvc-aircraftHPT#11Harmonic mean: 40.28
prompt-engineering-on-imagenetHPT#11Harmonic mean: 74.17
prompt-engineering-on-imagenet-rHPT#12Top-1 accuracy %: 77.38
prompt-engineering-on-oxford-iiit-pet-datasetHPT#12Harmonic mean: 96.71
prompt-engineering-on-dtdHPT#14Harmonic mean: 72.16
prompt-engineering-on-food-101HPT#14Harmonic mean: 91.01
prompt-engineering-on-imagenet-sHPT#14Top-1 accuracy %: 49.36
prompt-engineering-on-eurosatHPT#16Harmonic mean: 84.82
prompt-engineering-on-stanford-carsHPT#17Harmonic mean: 75.57