Computer Vision

CNN Robustness research. Application to face detectors and face ID systems
A short overview of Adversarial Robustness concentrated around the theoretic aspects.
Dec 1, 2020 12:00 AM Huawei Workshop on DL Theory
Aleksandr Petiushko Александр Петюшко
CNN Adversarial Robustness. The Theory (digital adversaries)
A post about the Convolution Neural Networks, their brittleness w.r.t. adversarial attacks in the digital domain, and the theory behind it.
Aleksandr Petiushko Александр Петюшко
Nov 8, 2020 1 min read
CNN Adversarial Robustness. The Theory (digital adversaries)
Anomaly detection for AI
An approach to tacke multiple problems for robustness, uncertainty, out-of-distribution, interpretability, and anomalies.
Nov 1, 2020 12:00 AM Huawei VI Summit and ML Workshop
Aleksandr Petiushko Александр Петюшко
Adversarial and Certified Robustness
Analysis of both empirical and theoretical approaches to study Adversarial Robustness (Adversarial Attacks, Certified Robustness)
Sep 20, 2020 12:00 AM ODS DataFest 2020
Aleksandr Petiushko Александр Петюшко
Slides
Adversarial and Certified Robustness
Black-Box Face Recovery from Identity Features
Reconstruction of face using only similarity and norm of the embedding of a blac-box FaceID system.
Anton Razzhigaev, Klim Kireev, Edgar Kaziakhmedov, Nurislam Tursynbek, Aleksandr Petiushko Александр Петюшко
PDF Cite Code Slides Video Source Document DOI arXiv WorkshopWeb
Black-Box Face Recovery from Identity Features
CNN Robustness Research. The Theory
A post about the Convolution Neural Networks, their brittleness w.r.t. adversarial attacks, and the theory behind it.
Aleksandr Petiushko Александр Петюшко
Aug 12, 2020 1 min read In Russian
CNN Robustness Research. The Theory
Application of Adversarial Machine Learning to Analyze the NN Robustness in the Real World
Effect of Adversarial ML on Real-World Robustness for Computer Vision (face detection and identification).
Jun 26, 2020 12:00 AM Spice IT Meetup
Aleksandr Petiushko Александр Петюшко
Slides Video In Russian
Application of Adversarial Machine Learning to Analyze the NN Robustness in the Real World
Adversarial Machine Learning in Real World
Analysis of Adversarial ML for Real-World Applications.
May 29, 2020 12:00 AM SPbSU Chebyshev Seminars
Aleksandr Petiushko Александр Петюшко
Slides Video In Russian
Adversarial Machine Learning in Real World
Certified Robustness for modern NN
The overview of certified robustness problems.
May 24, 2020 12:00 AM HSE/SkolTech, Statistical Learning Olympiad
Aleksandr Petiushko Александр Петюшко
Slides
Certified Robustness for modern NN
On Adversarial Patches: Real-World Attack on ArcFace-100 Face Recognition System
Two versions of a real-world adversarial attack (nose patch and adversarial glasses) for a top FaceID system - Arcface.
Mikhail Pautov, Grigorii Melnikov, Edgar Kaziakhmedov, Klim Kireev, Aleksandr Petiushko Александр Петюшко
PDF Cite Source Document DOI arXiv
On Adversarial Patches: Real-World Attack on ArcFace-100 Face Recognition System