Hierarchical-Prompting-for-Diffusion-Classifiers

Code release for "Hierarchical Prompting for Diffusion Classifiers" (ACCV 2024)

Installation

Create a conda environment with the following command:

conda env create -f environment.yml

Train

python train.py --dataset aircraft --n_trials 1 --to_keep 1 --n_samples 25 --loss l1 --prompt_path prompts/aircraft_name.csv

When applied to different datasets, the set up prompts part in the train.py needs to be modified accordingly, refer to the files in /prompts/xxx.csv.

Test

python test_hierarchical.py --dataset aircraft --n_trials 1  --to_keep 5 1 --n_samples 5 25  --prompt_path prompts/aircraft_name.csv

Citation