Awesome

Forbes (ECCV2024)

Jintae Kim, Seungwon Yang, Seong-Gyun Jeong, and Chang-Su Kim

Official code for "Forbes: Face Obfuscation Rendering via Backpropagation Refinement Scheme"[paper]

Requirements

• PyTorch 1.13.1
• CUDA 11.6
• python 3.8

Installation

Download repository:

$ git clone https://github.com/mcljtkim/Forbes.git

Create conda environment:

$ cd env
$ sh create_env.sh

Download AdaFace pre-trained model parameters from (https://github.com/mk-minchul/AdaFace).

Direct link to the parameters: pre-trained model.

Generate two folders and put the weights to the "weights" folder.

$ mkdir output
$ mkdir weights

Quick Usage

Generate an output image

$ python demo.py --img_path image_path/input_image.png

Citation

Please cite the following paper if you feel this repository useful.

@inproceedings{kim2024Forbes,
    author      = {Kim, Jintae and Yang, Seungwon and Jeong, Seong-Gyun and Kim, Chang-Su},
    title       = {Forbes: Face Obfuscation Rendering via Backpropagation Refinement Scheme},
    booktitle   = {Eur. Conf. Comput. Vis.},
    year        = {2024}
}