Tribhuvanesh Orekondy

I am a PhD student at the Max Planck Institute for Informatics where I work on Computer Vision, Machine Learning, and Privacy/Security. My advisors are Mario Fritz and Professor Bernt Schiele. Previously, I graduated with a Master's degree in CS from ETH Zürich.

Email  ·  Google Scholar  ·  Github  ·  LinkedIn  ·  MPI



I'm broadly interested in computer vision, machine learning, privacy, and security. Currently, my research focuses on the "double-edged" sword of ML. While I study ML models to understand and control privacy in data, I also address how these very ML models are susceptible to adversarial attacks.

Prediction Poisoning
Prediction Poisoning: Utility-Constrained Defenses Against Model Stealing Attacks
Tribhuvanesh Orekondy, Bernt Schiele Mario Fritz
arXiv preprint, 2019
paper  ·  bibtex

Perturbing posterior predictions by maximizing angular deviation of gradient signals results in reasonable trade-offs to defend against model stealing attacks.

Knockoff Nets: Stealing Functionality of Black-Box Models
Tribhuvanesh Orekondy, Bernt Schiele Mario Fritz
CVPR, 2019
paper  ·  poster  ·  extended abstract (CV-COPS@CVPR)  ·  project page  ·  bibtex

Functionality of complex blackbox vision models can be easily "knocked-off", even under minimal assumptions.


Understanding and Controlling User Linkability in Decentralized Learning
Tribhuvanesh Orekondy, Seong Joon Oh, Bernt Schiele Mario Fritz
arXiv preprint, 2018
paper  ·  bibtex

Gradient model updates in federated learning encodes non-IIDness of participating clients, raising linkability concerns.

Visual Redactions

Connecting Pixels to Privacy and Utility: Automatic Redaction of Private Information in Images
Tribhuvanesh Orekondy, Mario Fritz, Bernt Schiele
CVPR, 2018 (Spotlight)
paper  ·  poster  ·  project page  ·  video  ·  bibtex

Automatic method to identify and redact a broad range of private information spanning multiple modalities in visual content.


Towards a Visual Privacy Advisor: Understanding and Predicting Privacy Risks in Images
Tribhuvanesh Orekondy, Bernt Schiele, Mario Fritz
ICCV, 2017
paper  ·  poster  ·  extended abstract (VSM@ICCV)  ·  project page  ·  bibtex

An approach to understand and predict a wide spectrum of privacy risks in images.


HADES: Hierarchical Approximate Decoding for Structured Prediction
Tribhuvanesh Orekondy (under supervision of Martin Jaggi, Aurelien Lucchi, Thomas Hoffman )
Master Thesis, 2016
paper  ·  project page  ·  bibtex

A fast structured output learning algorithm, which works by approximately decoding oracles to various extents.

Academic Activities

  • PC member: CVPR '19, CV-COPS '19, ICCV '20
  • Teaching Assistant: Machine Learning in Cyber Security, 2018
  • Student supervision: Shadi Rahimian (MSc., University of Saarland), Jonas Klesen (BSc., University of Saarland)