search

ybendou / easy

This repository is the official implementation Ensemble Augmented-Shot Y-shaped Learning: State-Of-The-Art Few-Shot Classification with Simple Ingredients.
MIT License
112 stars 18 forks source link

Reproduce the outcome of 2x_ResNet12_√2 #15

Closed LIUZIJING-CHN closed 2 years ago

LIUZIJING-CHN commented 2 years ago

First of all, thx for your work! Since I want to reproduce the outcome of 2xResNet12√2 which has an accuracy of 70 in 1-shot case on mini-ImageNet, I want to know what is the training setting of the main.py? I want to extract the features from my own-trained model, hope you can help me.

ybendou commented 2 years ago

Hello, To do so, you will need to generate two sets of features (one for each ResNet12_√2) You can use the following 2 commands to get 2 features : 

python main.py --dataset-path <dataset-path> --dataset <dataset_name> --model resnet12 --feature-maps 45 --epochs 0  --batch-size 128 --n-shots 1  --load-model <your_model_path> --save-features <path_to_features1> --sample-aug 20
python main.py --dataset-path <dataset-path> --dataset <dataset_name> --model resnet12 --feature-maps 45 --epochs 0  --batch-size 128 --n-shots 1  --load-model <your_model_path> --save-features <path_to_features2> --sample-aug 20

Then finally you can test your results as follows:

python main.py --dataset-path "<dataset-path>" --dataset <dataset_name> --model resnet12 --test-features "[<path_to_features1>, <path_to_features2>]" --preprocessing ME --n-shots 1

Sorry for the delayed answer, I hope this helps. Let me know if it doesn't work.

Best,

LIUZIJING-CHN commented 2 years ago

I see. Thx for your reply!!