/2020bboxrestoration

Presentation for the paper "Black-Box Face Recovery from Identity Features"

Primary LanguageJupyter NotebookMIT LicenseMIT

Black-Box Face Recovery from Identity Features

Paper

The paper "Black-Box Face Recovery from Identity Features" by Anton Razzhigaev, Klim Kireev, Edgar Kaziakhmedov, Nurislam Tursynbek and Aleksandr Petiushko was accepted by ECCV-2020 Workshop "The Bright and Dark Sides of Computer Vision: Challenges and Opportunities for Privacy and Security" (CV-COPS 2020).

Paper on Arxiv.

Talk video

Youtube.

Presentation

Presentation.

Algorithm

requirements:

  • python 3.7
  • torch 1.5

Instructions:

Run all cells in

face_recovery.ipynb

Restoration dynamics

Restoration dynamics