search

Yikai-Wang / ICI-FSL

This repository contains the code for Instance Credibility Inference (ICI).
82 stars 22 forks source link

Hello, where are the csv files needed in code? #2

Closed Willforcv closed 4 years ago

Willforcv commented 4 years ago

thanks ~

Yikai-Wang commented 4 years ago

We suggest to create your own csv file to match the path in your machine. In the csv file, each line is related to one image where the first column is the path and the second column is the category/label. You may check/modify the code in: https://github.com/Yikai-Wang/ICI-FSL/blob/0d6a3e5b3403a8a3d8f22b04f91406fa4650fd97/datasets.py#L19-L35

Willforcv commented 4 years ago

We suggest to create your own csv file to match the path in your machine. In the csv file, each line is related to one image where the first column is the path and the second column is the category/label. You may check/modify the code in: https://github.com/Yikai-Wang/ICI-FSL/blob/0d6a3e5b3403a8a3d8f22b04f91406fa4650fd97/datasets.py#L19-L35

ok, thanks>!

songquanpeng commented 4 years ago

We suggest to create your own csv file to match the path in your machine. In the csv file, each line is related to one image where the first column is the path and the second column is the category/label. You may check/modify the code in: https://github.com/Yikai-Wang/ICI-FSL/blob/0d6a3e5b3403a8a3d8f22b04f91406fa4650fd97/datasets.py#L19-L35

Hello, @Yikai-Wang could you please upload a sample CSV file for us to reference? Really appreciate it!

Yikai-Wang commented 4 years ago

Hello, @Yikai-Wang could you please upload a sample CSV file for us to reference? Really appreciate it!

The csv file should be something like this (with two columns and several rows): filename label
n0153282900000005.jpg n01532829
n0153282900000006.jpg n01532829
n0153282900000007.jpg n01532829
n0153282900000010.jpg n01532829
songquanpeng commented 4 years ago

Hello, @Yikai-Wang could you please upload a sample CSV file for us to reference? Really appreciate it!

The csv file should be something like this (with two columns and several rows):

filename label n0153282900000005.jpg n01532829 n0153282900000006.jpg n01532829 n0153282900000007.jpg n01532829 n0153282900000010.jpg n01532829

Thanks for your reply. BTW, I wrote a simple script to help generate those specified CSV files: generate_required_csv.py, hope it helpful for other guys.