Prof. Dr. Mario Fritz Faculty | ![]() |
---|
My group is working on Trustworthy Information Processing with a focus on the intersection of AI & Machine Learning with Security & Privacy.
Recent publications:
- USENIX Security’20: Updates-Leak: Data Set Inference and Reconstruction Attacks in Online Learning
- TPAMI’19: Person Recognition in Personal Photo Collections
- Explainable AI Book: Towards reverse-engineering black-box neural networks
- PriML@NeurIPS’19: Differential Privacy Defenses and Sampling Attacks for Membership Inference
- PriML@NeurIPS’19: GAN-Leaks: A Taxonomy of Membership Inference Attacks against GANs
- FL@NeurIPS’19: Gradient-Leaks: Understanding Deanonymization in Federated Learning
- BDL@NeurIPS’19: Conditional Flow Variational Autoencoders for Structured Sequence Prediction
- BDL@NeurIPS’19: “Best of Many” Samples Distribution Matching
- HotPETs’19: Updates-Leak: Data Set Inference and Reconstruction Attacks in Online Learning
- PUT@PETs’19: Understanding and Recognizing Bystanders in Images for Privacy Protection
- ICCV’19: Attributing Fake Images to GANs: Learning and Analyzing GAN Fingerprints
- ICCV’19: Deep Appearance Maps
- CVPR’19: Knockoff Nets: Stealing Functionality of Black-Box Models
- CVPR’19: Not Using the Car to See the Sidewalk: Quantifying and Controlling the Effects of Context in Classification and Segmentation
- CVPR’19: Time-Conditioned Action Anticipation in One Shot
- ICLR’19: Bayesian Prediction of Future Street Scenes using Synthetic Likelihoods
- NDSS’19: ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models
- WACV’19: Fashion is Taking Shape: Understanding Clothing Preference Based on Body Shape From Online Sources
- TPAMI’19: MPIIGaze: Real-World Dataset and Deep Appearance-Based Gaze Estimation
- NIPS’18: Adversarial Scene Editing: Automatic Object Removal from Weak Supervision
- ECCV’18: A Hybrid Model for Identity Obfuscation by Face Replacement
- ECCV’18: Diverse Conditional Image Generation by Stochastic Regression with Latent Drop-Out Codes
- ECCV-W’18: Answering Visual What-If Questions: From Actions to Predicted Scene Descriptions
- CSCS’18: Sequential Attacks on Agents for Long-Term Adversarial Goals
- USENIX Security’18: A4NT: Author Attribute Anonymity by Adversarial Training of Neural Networks
- CVPR’18: Connecting Pixels to Privacy and Utility: Automatic Redaction of Private Information in Images
- CVPR’18: Natural and Effective Obfuscation by Head Inpainting
- CVPR’18: Disentangled Person Image Generation
- CVPR’18: Long-Term On-Board Prediction of People in Traffic Scenes under Uncertainty
- CVPR’18: Accurate and Diverse Sampling of Sequences based on a “Best of Many” Sample Objective
- ICLR’18: Towards Reverse-Engineering Black-Box Neural Networks
- Scientific Reports’18: Advanced Steel Microstructural Classification by Deep Learning Methods
- AAAI’18: Long-Term Image Boundary Prediction
- TPAMI’18: Reflectance and Natural Illumination from Single-Material Specular Objects Using Deep Learning
Most recent work on ArXiv:
- ArXiv’19: “Best-of-Many-Samples” Distribution Matching
- ArXiv’19: WhiteNet: Phishing Website Detection by Visual Whitelists
- ArXiv’19: Interpretability Beyond Classification Output: Semantic Bottleneck Networks
- ArXiv’19: Prediction Poisoning: Utility-Constrained Defenses Against Model Stealing Attacks
- ArXiv’19: SampleFix: Learning to Correct Programs by Sampling Diverse Fixes
- ArXiv’19: Shape Evasion: Preventing Body Shape Inference of Multi-Stage Approaches
- ArXiv’19: Learning Manipulation under Physics Constraints with Visual Perception
- ArXiv’18: MLCapsule: Guarded Offline Deployment of Machine Learning as a Service
- ArXiv’18: Understanding and Controlling User Linkability in Decentralized Learning
News, talks, events:
- Keynote at Workshop Machine Learning for Cybersecurity, ECMLPKDD’19
- Talk at Cyber Defense Campus (CYD) Conference on Artificial Intelligence in Defence and Security
- Co-Organizer of Second International Workshop on The Bright and Dark Sides of Computer Vision: Challenges and Opportunities for Privacy and Security (CV-COPS) at CVPR 2019
- Co-Organizer: 3. ACM Symposium on Computer Science in Cars: Future Challenges in Artificial Intelligence & Security for Autonomous Vehicles CSCS’19
- Leading scientist at new Helmholtz Medical Security and Privacy Research Center
- Member of ACM Technical Policy Committee Europe
- Mateusz Malinowski received the DAGM MVTec dissertation award as well as the Dr.-Eduard-Martin award for his PhD
- Associate Editor for IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
- Talk on “Challenges of Disruptive Technologies (BigData/AI) for MINT Education” at “MINT Conference: MINT Education towards 2028“
- Talk at “Sommer der Cybersicherheit“, Congresshalle, Saarbrücken
- Co-Organizer: new Conference on Future Challenges in
Artificial Intelligence & Security for Autonomous Vehicles CSCS’18 - Co-Organizer of Workshop on Interactive and Adaptive Learning in an Open World at ECCV’18
- Co-Organizer of Second International Workshop on The Bright and Dark Sides of Computer Vision:
Challenges and Opportunities for Privacy and Security (CV-COPS) at CVPR 2018 - Invited talk Center for Art and Media, Karlsruhe, 2018
“The Bright and Dark Side of Computer Vision: Latest Advances and Implications on Privacy” - Program Chair of German Conference on Pattern Recognition GCPR 2018
- Invited talk at First International Workshop on The Bright and Dark Sides of Computer Vision:
Challenges and Opportunities for Privacy and Security at CVPR 2017
“Towards a Visual Privacy Advisor: Understanding and Controlling Privacy Risks in Visual Data” - Invited talk at ACM Chapters Computer Science in Cars Symposium, 2017
- Invited talk at Symposium on Image Forensics and Identification, 2017
“Re-Identification with Deep Learning”
2014 |
|
Technical Reports |
|
![]() | Ubic: Bridging the gap between digital cryptography and the physical world Technical Report arXiv:1403.1343 [cs.CR], 2014. |