2024

Conferences

  • M. Augustin, Y. Neuhaus, M. Hein: Analyzing and Explaining Image Classifiers via Diffusion Guidance, CVPR 2024, PDF
  • F. Croce, M. Hein. Segment (Almost) Nothing: Prompt-Agnostic Adversarial Attacks on Segmentation Models. SaTML 2024. PDF

Workhops and Preprints

  • C. Schlarmann, N. D. Singh, F. Croce, M. Hein: Robust CLIP: Unsupervised Adversarial Fine-Tuning of Vision Embeddings for Robust Large Vision-Language Models,  Link arXiv:2402.12336, Github (accepted at ICLR 2024 Workshop on Reliable and Responsible Foundation Models and the ICLR 2024 on Mathematical and Empirical Understanding of Foundation Models)
  • M. Müller, M. Hein: How to train your ViT for OOD detection, ICLR 2024 Workshop on Reliable and Responsible Foundation Models

2023

Journals

  • A. Gautier, F. Tudisco, M. Hein (2023). Nonlinear Perron--Frobenius Theorems for Nonnegative Tensors. SIAM Review, 65(2), 495-536.
  • A. Topaldi, U. Bharadwaj, K. T. Gao, R. Bhattacharajee, F. G. Gassert, J. Luitjens, P. Giesler, J. N. Morshuis, P. Fischer, M. Hein, C. Baumgartner, A. Razumov, D. Dylov, Q. van Lohuizen, S. J. Fransen, X. Zhang, R. Tibrewala, H. L. de Moura, K. Liu, M. V. W. Zibetti, R. Regatte, S. Majumdar, V. Pedoi (2023). K2S Challenge: From Undersampled K-Space to Automatic Segmentation. Bioengineering, 10(2), 267 (Special Issue: AI in MRI, Frontiers and Applications). Link

Conferences

  • N. D. Singh, F. Croce, M. Hein. Revisiting Adversarial Training for ImageNet: Architectures, Training and Generalization across Threat Models. NeurIPS 2023. PDF (Models available on RobustBench and here)
  • M. Müller, T. J. Vlaar, D. Rolnick, M. Hein. Normalization Layers are All that Sharpness-Aware Minimization needs. NeurIPS 2023. PDF Code
  • Y. Neuhaus, M. Augustin, V. Boreiko, M. Hein. Spurious Features Everywhere - Large-Scale Detection of Harmful Spurious Features in ImageNet. ICCV 2023. PDF Code/Dataset
  • J. Bitterwolf, M. Mülller, M. Hein. In or Out? Fixing ImageNet Out-of-Distribution Detection Evaluation (2023). ICML 2023.  PDF Code/Dataset
  • M. Andriushchenko, F. Croce, M. Müller, M. Hein, N. Flammarion. A modern look at the relationship between sharpness and generalization.  ICML 2023 PDF Code
  • V. Voráček, M. Hein (2023). Improving l1-certified randomized smoothing by leveraging box constraints. ICML 2023. PDF Code
  • M. Yatsura, K. Sakmann, N. Grace Hua, M. Hein, J.-H. Metzen. Certified Defences Against Adversarial Patch Attacks on Semantic Segmentation. ICLR 2023. PDF
  • V. Voráček, M. Hein (2023). Sound randomized smoothing in floating-point arithmetics. ICLR 2023. PDF   Code

Workshops and Preprints

  • M. Augustin, Y. Neuhaus, M. Hein (2023). Analyzing and Explaining Image Classifiers via Diffusion Guidance. PDF
  • I.Ilanchezian, V. Boreiko, L. Kühlewein, Z. Huang, M. Ayhan, M. Hein, L. Koch, P. Berens (2023). Generating Realistic Counterfactuals for Retinal Fundus and OCT Images using Diffusion Models. PDF
  • V. Boreiko, M. Hein, J.-H. Metzen (2023). Identifying Systematic Errors in Object Detectors with the SCROD Pipeline.  ICCV Workshop "Robustness and Reliability of Autonomous Vehicles in the Open-world. PDF
  • C. Schlarmann, M. Hein (2023). On the Adversarial Robustness of Multi-Modal Foundation Models. ICCV Workshop on Adversarial Robustness in the Wild Link PDF
  • F. Croce, N. D. Singh, M. Hein (2023). Robust Semantic Segmentation: Strong Adversarial Attacks and Fast Training of Robust Models. arXiv PDF, ICML Workshop New Frontiers in Adversarial Machine Learning.
  • J. Bitterwolf, M. Mülller, M. Hein (2023). In or Out? Fixing ImageNet Out-of-Distribution Detection Evaluation. ICLR 2023 Workshop on Trustworthy Machine Learning PDF/ ICLR 2023 Trustworthy and Reliable Large-Scale Machine Learning Models PDF/ CVPR 2023 Workshop VAND: Visual Anomaly and Novelty Detection/ ICML 2023 Workshop on Data-Centric Machine Learning. Code/Dataset

2022

Journals

  • D. Stutz, N. Chandramoorthy, M. Hein, B. Schiele (2022). Random and Adversarial Bit Error Robustness: Energy-Efficient and Secure DNN Accelerators. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI). PDF Project

Conferences

  • M. Augustin*, V. Boreiko*, F. Croce, M. Hein (2022). Diffusion Visual Counterfactual Explanations. NeurIPS 2022 (* joint first author). PDF Code
  • A. Meinke, J. Bitterwolf, M. Hein (2022). Provably Adversarially Robust Detection of Out-of-Distribution Data (Almost) for Free. NeurIPS 2022. PDF Code
  • V. Boreiko, M. Augustin, F. Croce, P. Berens, M. Hein (2022). Sparse Visual Counterfactual Explanations in Image Space. GCPR 2022. PDF Code
  • V. Boreiko, I. Ilanchezian, M. Ayhan, S. Müller, L. Koch, P. Berens, M. Hein (2022). Visual counterfactual explanations for robust disease detection in ophthalmology. MICCAI 2022, PDF
  • V. Voráček, M. Hein (2022). Provably Adversarially Robust Nearest Prototype Classifiers. ICML 2022, PDF Code
  • F. Croce, M. Hein (2022). Adversarial robustness against multiple and single l_p-threat models via quick fine-tuning of robust classifiers. ICML 2022, PDF Code
  • F. Croce, S. Gowal, T. Brunner, E. Shelhamer, M. Hein, T. Cemgil (2022). Evaluating the Adversarial Robustness of Test-Time Defenses. ICML 2022, PDF Code
  • J. Bitterwolf, A. Meinke, M. Augustin, M. Hein (2022). Breaking Down Out-of-Distribution Detection: Many Methods Based on OOD Training Data Estimate a Combination of the Same Core Quantities. ICML 2022, PDF Code
  • D. Heller, P. Ferber, J. Bitterwolf, M. Hein, J. Hoffmann (2022). Neural Network Heuristic Functions: Taking Confidence into Account, Symposium on Combinatorial Search. Symposium on Combinatorial Search (SOCS 2022).
  • A. Kristiadi, M. Hein, P. Hennig (2022). Being a Bit Frequentist Improves Bayesian Neural Networks. AISTATS 2022,  PDF Code
  • F. Croce, M. Andriushchenko, N. Singh, N. Flammarion, M. Hein (2022). Sparse-RS: a versatile framework for query-efficient sparse black-box adversarial attacks. AAAI 2022, PDF Code

Workshops

  • M. Yatsura, K. Sakmann, N. Grace Hua, M. Hein, J.-H. Metzen (2022). Certified Defences Against Adversarial Patch Attacks on Semantic Segmentation. NeurIPS 2022 Workshop on Trustworthy and Socially Responsible Machine Learning. PDF
  • M. Müller, M. Hein (2022). Perturbing BatchNorm and Only BatchNorm Benefits Sharpness-Aware Minimization. NeurIPS 2022 Workshop ``Has it Trained Yet? Algorithmic Efficiency in Practical Neural Network Training''. PDF
  • N. Morshius, S. Gatidis, M. Hein, C. Baumgartner (2022). Adversarial Robustness of MR Image Reconstruction under Realistic Perturbations. MICCAI Workshop on Machine Learning for Medical Image Reconstruction. PDF Code
  • V. Voráček, M. Hein (2022). Sound randomized smoothing in floating-point arithmetics. ICML 2022 Workshop on Formal Verification of Machine Learning and ICML 2022 Workshop on Adversarial Machine Learning Frontiers. PDF (long version) Code
  • F. Croce and M. Hein (2022). On the interplay of adversarial robustness and architecture components: patches, convolution and attention. ICML 2022 Workshop on Adversarial Machine Learning Frontiers. PDF
  • A. Meinke, J. Bitterwolf, M. Hein (2022). Provably Robust Detection of Out-of-distribution Data (almost) for free. ICML 2022 Workshop on Adversarial Machine Learning Frontiers.
  • V. Boreiko, M. Augustin, F. Croce, P. Berens, M. Hein (2022). Sparse Visual Counterfactual Explanations in Image Space, CVPR 2022 Workshop on The Art of Robustness: Devil and Angel in Machine Learning. PDF Code

2021

Journals

  • A. Gautier, M. Hein, F. Tudisco (2021). The global convergence of the nonlinear power method for mixed-subordinate matrix norms. Journal of Scientific Computing, 88, DOI 10.1007/s10915-021-01524-w.  PDF

Conferences

  • F. Croce, M. Andriushchenko, V. Sehwag, N. Flammarion, M. Chiang, P. Mittal, M. Hein (2021). RobustBench: a standardized adversarial robustness benchmark. NeurIPS 2021 Dataset and Benchmark Track  PDF  Link
  • M. Yatsura, J. Metzen, M. Hein (2021). Meta-Learning the Search Distribution of Black-Box Random Search Based Adversarial Attacks. NeurIPS 2021. PDF Code
  • A. Kristiadi, M. Hein, P. Hennig (2021). An Infinite-Feature Extension for Bayesian ReLU Nets That Fixes Their Asymptotic Overconfidence. NeurIPS 2021 (also presented at ICLR 2021 Workshop on Robust and Reliable Machine Learning in the Real World)PDF
  • D. Stutz, M. Hein, B. Schiele (2021). Relating Adversarially Robust Generalization to Flat Minima. oral at ICCV 2021 (also presented at CVPR 2021 Workshop on Adversarial Machine Learning in Real-World Computer Vision Systems). PDF Project
  • A. Kristiadi, M. Hein, P. Hennig (2021). Learnable Uncertainty under Laplace Approximations. UAI 2021. PDF Code
  • F. Croce, M. Hein (2021). Mind the box: l_1-APGD for sparse adversarial attacks on image classifiers. ICML 2021 (also presented at ICLR 2021 Workshop on Security and Safety of Machine Learning Systems) PDF Code
  • D. Stutz, N. Chandramoorthy, M. Hein, B. Schiele (2021). Bit Error Robustness for Energy-Efficient DNN Accelerators. MLSys 2021 (also presented at ICLR 2021 Workshop on Robust and Reliable Machine Learning in the Real World and CVPR 2021 Workshop on Adversarial Machine Learning in Real-World Computer Vision Systems)  PDF Code

Workshops and Preprints

  • A. Kristiadi, M. Hein, P. Hennig (2021). Being a Bit Frequentist Improves Bayesian Neural Networks. NeurIPS 2021 Workshop on Bayesian Deep Learning   PDF
  • A. Meinke, J. Bitterwolf, M. Hein (2021). Provably Robust Detection of Out-of-distribution Data (almost) for free. contributed talk at ICML 2021 Workshop on Uncertainty & Robustness in Deep Learning PDF Code
  • F. Croce, M. Hein (2021). Adversarial robustness against multiple l_p-threat models at the price of one and how to quickly fine-tune robust models to another threat model. preprint. PDF Code
  • F. Croce, M. Andriushchenko, V. Sehwag, N. Flammarion, M. Chiang, P. Mittal, M. Hein (2021). RobustBench: a standardized adversarial robustness benchmark. ICLR Workshop on Security and Safety of Machine Learning Systems PDF (Best Paper Honorable Mention Prize) Link
  • M. Augustin, M. Hein (2021): Out-distribution aware Self-training in an Open World Setting.  available on arxiv PDF

2020

Journals

  • N. Garcia Trillos, M. Gerlach, M. Hein, D. Slepcev (2020): Error estimates for spectral convergence of the graph Laplacian on random geometric graphs towards the Laplace-Beltrami operator. Foundations of Computational Mathematics,  20, 827–887. PDF  (Link to journal)

Conferences

  • J. Bitterwolf, A. Meinke, M. Hein (2020): Certifiably Adversarially Robust Detection of Out-of-Distribution Data. NeurIPS 2020 PDF, presented also at ICML 2020 Workshop on Uncertainty and Robustness   Code
  • M. Augustin, A. Meinke, M. Hein (2020): Adversarial Robustness on In- and Out-Distribution Improves Explainability.  ECCV 2020 PDF Code
  • M. Andriushchenko, F. Croce, N Flammarion, M. Hein (2020): Square Attack: a query-efficient black-box adversarial attack via random search.  ECCV 2020 PDF   Code
  • F. Croce, M. Hein (2020): Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks. ICML 2020 PDF    Benchmark and Code  (also spotlight at the ICML 2020 Workshop on Uncertainty and Robustness)
  • A. Kristiadi, M. Hein, P. Hennig (2020): Being Bayesian, Even Just a Bit, Fixes Overconfidence in ReLU Networks. ICML 2020 PDF
  • D. Stutz, M. Hein, B. Schiele (2020):  Confidence-Calibrated Adversarial Training: Generalizing to Unseen Attacks. ICML 2020  PDF    Code (also oral at the ICML Workshop on Uncertainty and Robustness)
  • F. Croce, M. Hein (2020): Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack. ICML 2020 PDF Code
  • A. Meinke, M. Hein (2020): Towards neural networks that provably know when they don't know. ICLR 2020  PDF    Code
  • F. Croce, M. Hein (2020): Provable robustness against all adversarial l_p-perturbations for p>=1.  ICLR 2020 PDF   Code

Workshops and Preprints

  • D. Stutz, N. Chandramoorthy, M. Hein, B. Schiele (2020). Bit Error Robustness for Energy-Efficient DNN Accelerators. preprint PDF Code
  • F. Croce, M. Andriushchenko, N. D. Singh, N. Flammarion, M. Hein: Sparse-RS: a versatile framework for query-efficient sparse black-box adversarial attacks. preprint PDF, to be presented at ECCV 2020 Workshop Adversarial Robustness in the Real World, Code
  • A. Gautier, M. Hein, F. Tudisco (2020): Computing the norm of nonnegative matrices and the log-Sobolev constant of Markov chains. preprint PDF

2019

Journals

  • A. Gautier, F. Tudisco, M. Hein (2019): A unifying Perron-Frobenius theorem for nonnegative tensors via multi-homogenous maps. SIAM J. Matrix Analysis (SIMAX), Vol. 40, No. 3, pp. 1206–1231. PDF
  • A. Gautier, F. Tudisco, M. Hein (2019): The Perron-Frobenius theorem for multi-homogeneous mappings. SIAM J. Matrix Analysis (SIMAX), Vol. 40, No. 3, pp. 1179--1205. PDF
  • F. Croce, J. Rauber, M. Hein (2019): Scaling up the randomized gradient-free adversarial attack reveals overestimation of robustness using established attacks. Accepted at International Journal of Computer Vision (IJCV). PDF (Link to journal) Code

Conferences

  • M. Andriushchenko, F. Croce, N Flammarion, M. Hein (2019): Square Attack: a query-efficient black-box adversarial attack via random search. preprint PDF   Code
  • M. Andriushchenko, M. Hein (2019): Provably Robust Boosted Decision Stumps and Trees against Adversarial Attacks. NeurIPS 2019 PDF (oral presentation in NeurIPS 2019 Workshop "Machine Learning with Guarantees")    Code
  • P. Mercado, F. Tudisco, M. Hein (2019): Generalized Matrix Means for Semi-Supervised Learning with Multilayer Graphs. NeurIPS 2019. PDF (long versionCode
  • F. Croce, M. Hein (2019): Sparse and imperceivable adversarial attacks. ICCV 2019. PDF  Code
  • M. Hein, M. Andriushchenko, J. Bitterwolf (2019): Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem. CVPR 2019 (oral presentation) - (also oral at ICML 2019 Workshop on  Uncertainty & Robustness in Deep Learning). PDF   Code
  • D. Stutz, M. Hein, B. Schiele (2019): Disentangling Adversarial Robustness and Generalization. CVPR 2019. PDF   Code
  • F. Croce, M. Andriushchenko, M. Hein (2019): Provable Robustness of ReLU networks via Maximization of Linear Regions. AISTATS 2019. PDF     Code
  • Q. Nguyen, M. Mukkamala, M. Hein (2019): On the loss landscape of a class of deep neural networks with no bad local valleys. ICLR 2019. PDF

Workshops and Preprints

  • F. Croce, M. Hein (2019): Provable robustness against all adversarial l_p-perturbations for p>=1. NeurIPS 2019 Workshop "Machine Learning with Guarantees" PDF   Code
  • A. Meinke, M. Hein (2019): Towards neural networks that provably know when they don't know. NeurIPS 2019 Workshop "Machine Learning with Guarantees"  PDF   Code
  • F. Croce, M. Hein (2019): Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack. preprint PDF Code
  • P. Mercado, F. Tudisco, M. Hein (2019): Spectral Clustering of Signed Graphs via Matrix Power Means. ICML 2019. PDF  Code

2018

  • ​​​​​​M. Mosbach, M. Andriushchenko, T. Trost, M. Hein, D. Klakow (2018): Logit Pairing Methods Can Fool Gradient-Based Attacks. NeurIPS 2018 Workshop on Security in Machine Learning. PDF  Code
  • F. Croce, M. Hein (2018): A randomized gradient-free attack on ReLU networks. GCPR 2018.  PDF  Code
  • F. Tudisco, P. Mercado, M. Hein (2018): Community Detection in Networks via Nonlinear Modularity Eigenvectors. SIAM Journal of Applied Mathematics, 78(5): 2393--2419. PDF
  • M. Lapin, M. Hein, B. Schiele (2018): Analysis and Optimization of Loss Functions for Multiclass, Top-k, and Multilabel Classification. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 40(7):1533-1554. PDF
  • F. Tudisco, M. Hein (2018): A nodal domain theorem and a higher-order Cheeger inequality for the graph p-Laplacian. Journal of Spectral Theory, 8(3): 883-908. PDF
  • Q. Nguyen, M. Hein (2018): Optimization Landscape and Expressivity of Deep CNNs. ICML 2018. PDF (long version including proofs)
  • Q. Nguyen, M. Mukkamala, M. Hein (2018): Neural networks should be wide enough to learn disconnected decision regions. ICML 2018. PDF (long version including proofs)
  • N. Garcia Trillos, M. Gerlach, M. Hein, D. Slepcev (2018): Error estimates for spectral convergence of the graph Laplacian on random geometric graphs towards the Laplace--Beltrami operator. PDF
  • A. Gautier, F. Tudisco, M. Hein (2018): A unifying Perron-Frobenius theorem for nonnegative tensors via multi-homogeneous maps. PDF
  • A. Gautier, F. Tudisco, M. Hein (2018): The Perron-Frobenius theorem for multi-homogeneous mappings. PDF
  • P. Mercado, A. Gautier, F. Tudisco, M. Hein (2018): The Power Mean Laplacian for Multilayer Graph Clustering. AISTATS 2018. PDF (Appendix containing proofs) Code

2017

  • Q. Nguyen, M. Hein (2017): The loss surface and expressivity of deep convolutional neural networks. PDF
  • M. Hein, M. Andriushchenko (2017): Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation. NeurIPS 2017. PDF (long version including proofs) Code
  • M. C. Mukkamala, M. Hein (2017): Variants of RMSProp and Adagrad with Logarithmic Regret Bounds. ICML 2017. PDF (long version including proofs) Code
  • Q. Nguyen, M. Hein (2017): The loss surface of deep and wide neural networks. ICML 2017. PDF (long version including proofs)
  • A. Khoreva, R. Beneson, J. Hosang, M. Hein and B. Schiele (2017): Simple does it: Weakly Supervised Instance and Semantic Segmentation. CVPR 2017. PDF
  • A. Gautier, F. Tudisco, M. Hein (2017): The Perron-Frobenius Theorem for Multi-homogeneous Maps. PDF
  • P. Lutsik, M. Slawski, G. Gasparoni, N. Vedeneev, M. Hein, J. Walter (2017): MeDeCom: discovery and quantification of latent components of heterogeneous methylomes. Genome Biology, 18:55. PDF Code

2016

  • M. Lapin, M. Hein, B. Schiele (2016): Analysis and Optimization of Loss Functions for Multiclass, Top-k, and Multilabel Classification. PDF  Code
  • A. Gautier, Q. Nguyen Ngoc, M. Hein (2016): Globally Optimal Training of Generalized Polynomial Neural Networks with Nonlinear Spectral Methods. NeurIPS 2016. PDF (long version) Matlab Code
  • P. Mercado, F. Tudisco, M. Hein (2016): Clustering Signed Networks with the Geometric Mean of Laplacians. NeurIPS 2016. PDF (long version) Code
  • A. Gautier, M. Hein (2016): Tensor norm and maximal singular vectors of non-negative tensors - a Perron-Frobenius theorem, a Collatz-Wielandt characterization and a generalized power method. Linear Algebra and its Applications, 505:313–343. Article Preprint 
  • F. Tudisco, M. Hein (2016): A nodal domain theorem and a higher-order Cheeger inequality for the graph p-Laplacian. PDF
  • Q. Nguyen, F. Tudisco, A. Gautier, M. Hein (2016): An Efficient Multilinear Optimization Framework for Hypergraph Matching. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 39(6):1054-1075. PDF Code available on request
  • M. Lapin, M. Hein and B. Schiele (2016): Loss Functions for Top-k Error: Analysis and Insights. CVPR 2016. PDF (long version) Code
  • Y. Xian, Z. Akata, G. Sharma, Q. Ngoc, M. Hein, B. Schiele (2016): Latent Embeddings for Zero-shot Classification. CVPR 2016 (spotlight). PDF (long version
  • A. Khoreva, R. Beneson, M. Omran, M. Hein,  B. Schiele (2016): Weakly Supervised Object Boundaries. CVPR 2016 (spotlight). PDF (long version)

2015

  • M. Lapin, M. Hein, B. Schiele (2015): Top-k Multiclass SVM. Spotlight at NeurIPS 2015. PDF (long version) Code
  • P. Jawanpuria, M. Lapin, M. Hein, B. Schiele (2015): Efficient Output Kernel Learning for Multiple Tasks. NeurIPS 2015. PDF (long version) Code
  • M. Slawski, P. Li, M. Hein (2015): Regularization-Free Estimation in Trace Regression with Symmetric Positive Semidefinite Matrices. NeurIPS 2015. PDF (long version)
  • A. Gautier, M. Hein (2015): Tensor norm and maximal singular vectors of non-negative tensors - a Perron-Frobenius theorem, a Collatz-Wielandt characterization and a generalized power method. PDF 
  • Q. Ngoc, A. Gautier, M. Hein (2015): A Flexible Tensor Block Coordinate Ascent Scheme for Hypergraph Matching. CVPR 2015 (oral presentation). PDF (Supplementary material
  • A. Khoreva, F. Galasso, M. Hein, B. Schiele (2015): Classifier Based Graph Construction for Video Segmentation. CVPR 2015. PDF (Supplementary material
  • S. Bhadra, M. Hein (2015): Correction of Noisy Labels via Mutual Consistency Check. Neurocomputing, 160: 34-52. PDF LND CVScript 
  • M. Slawski, M. Hein (2015): Estimation of positive definite M-matrices and structure learning for attractive Gaussian Markov Random fields. Linear Algebra and its Applications, 473: 145-179 (Special Issue on Statistics). PDF

2014

  • S. Rangapuram, P. K. Mudrakarta, M. Hein (2014): Tight continuous relaxation of the balanced k-cut problem. NeurIPS 2014. PDF (long version) Code
  • L. Jost, S. Setzer, M. Hein (2014): Nonlinear Eigenproblems in Data Analysis - Balanced Graph Cuts and the RatioDCA-Prox. In: S. Dahlke, W. Dahmen, M. Griebel, W. Hackbusch, K. Ritter, R. Schneider, C. Schwab, H. Yserentant (Eds.): Extraction of Quantifiable Information from Complex Systems. Springer. PDF
  • A. Podosinnikova, S. Setzer, M. Hein (2014): Robust PCA: Optimization of the Robust Reconstruction Error on the Stiefel Manifold. GCPR 2014. PDF (Supplementary material) Code 
  • A. Khoreva, F. Galasso, M. Hein, B. Schiele (2014): Learning Must-Link Constraints for Video Segmentation based on Spectral Clustering. GCPR 2014. PDF 
  • M. Slawski, M. Hein (2014): Sparse Recovery for Protein Mass Spectrometry Data. In: I. Rish, G. Cecchi, A. Lozano, A. Niculescu-Mizil (Eds.): Practical Applications of Sparse Modeling. MIT press. PDF
  • U. von Luxburg, A. Radl, M. Hein (2014): Hitting and Commute Times in Large Random Neighborhood Graphs. Journal of Machine Learning Research, 15: 1751-1798. PDF
  • M. Lapin, M. Hein, B. Schiele (2014): Scalable Multitask Representation Learning for Scene Classification. CVPR 2014. PDF  Code
  • M. Lapin, M. Hein, B. Schiele (2014): Learning Using Privileged Information: SVM+ and Weighted SVM. Neural Networks, 53: 95-108. PDF

2013

  • M. Slawski, M. Hein (2013): Non-negative least squares for high-dimensional linear models: consistency and sparse recovery without regularization. Electronic Journal of Statistics, 7(0):3004-3056. PDF
  • M. Hein, S. Setzer, L. Jost, S. Rangapuram (2013): The Total Variation on Hypergraphs - Learning on Hypergraphs Revisited. Spotlight at NeurIPS 2013 (acceptance rate < 5%). PDF (long version)
  • M. Slawski, M. Hein, P. Lutsik (2013): Matrix Factorization with Binary Components. Spotlight at NeurIPS 2013 (acceptance rate < 5%). PDF (Supplementary material)
  • S. Rangapuram, T. Buehler, M. Hein (2013): Towards Realistic Team Formation in Social Networks based on Densest Subgraphs. WWW 2013, 1077-1088. PDF
  • T. Buehler, S. Rangapuram, S. Setzer, M. Hein (2013): Constrained fractional set programs and their application in local clustering and community detection. ICML 2013, 624-632. PDF (long version) Code
  • M. Maier, U. von Luxburg, M. Hein (2013): How the result of graph clustering methods depends on the construction of the graph. ESAIM: Probability and Statistics, 17: 370-418. PDF

2012

  • M. Slawski, R. Hussong, A. Tholey, T. Jakoby, B. Gregorius, A. Hildebrandt, M. Hein (2012): Isotope pattern deconvolution for peptide mass spectrometry by non-negative least squares/least absolute deviation template matching. BMC Bioinformatics 2012, 13:291 (8 November 2012). PDF
  • C. Backes, A. Rurainski, G.W. Klau, O. Müller, D. Stöckel, A. Gerasch, J. Küntzer, D. Maisel, N. Ludwig, M. Hein, A. Keller, H. Burtscher, M. Kaufmann, E. Meese, H.-P. Lenhof (2012): An integer linear programming approach for finding deregulated subgraphs in regulatory networks. Nucleic Acids Research, 40(6):e43. PDF
  • S. Rangapuram, M. Hein (2012): Constrained 1-Spectral Clustering. AISTATS 2012, JMLR W&CP 22: 1143-1151. PDF (Supplementary material)

2011

  • M. Slawski,M. Hein (2011): Sparse recovery by thresholded non-negative least squares. In: Advances in Neural Information Processing Systems 24 (NeurIPS 2011). 1926--1934. PDF (Supplementary material)
  • M. Hein, S. Setzer (2011): Beyond Spectral Clustering - Tight Relaxations of Balanced Graph Cuts. In: Advances in Neural Information Processing Systems 24 (NeurIPS 2011). 2366--2374. PDF (long version)
  • M. Slawski, M. Hein (2011): Robust sparse recovery with non-negativity constraints. In: Workshop on Signal Processing with Adaptive Sparse Structured Representations (SPARS). PDF

2010

  • M. Slawski, M. Hein (2010): Sparse Recovery for Protein Mass Spectrometry Data. In: NeurIPS Workshop: Practical Application of Sparse Modeling: Open Issues and New Directions. PDF
  • M. Hein, T. Buehler (2010): An inverse power method for nonlinear eigenproblems with applications in 1-spectral clustering and sparse PCA. In: Advances in Neural Information Processing Systems 23 (NeurIPS 2010). 847-855. PDF (long version) Code (1-spectral clustering)Code (sparse PCA)
  • U. von Luxburg, A. Radl, M. Hein (2010): Getting lost in space: Large sample analysis of the commute distance. In: Advances in Neural Information Processing Systems 23 (NeurIPS 2010). 2622-2630. PDF (long version)
  • F. Steinke, M. Hein, B. Schoelkopf )2010): Non-parametric regression between general Riemannian manifolds. SIAM Journal on Imaging Sciences, 3:527-563. PDF
  • U. von Luxburg, A. Radl, M. Hein (2010): Hitting times, commute distances and the spectral gap for large random geometric graphs. arXiv:1003.1266v1. PDF
  • M. Slawski, W. zu Castell, G. Tutz (2010): Feature Selection Guided by Structural Information. Annals of Applied Statistics, 4:1056-1080. PDF

2009

  • M. Hein (2009): Robust Nonparametric Regression with Metric-Space valued Output. In: Y. Bengio and D. Schuurmans, J. Lafferty, C. K. I. Williams, A. Culotta (Eds.) Advances in Neural Information Processing Systems 22 (NeurIPS 2009), 718-726, MIT Press, Cambridge, MA, 2010. PDF (long version)
  • K.I. Kim, F. Steinke, M. Hein (2009): Semi-supervised Regression using Hessian energy with an application to semi-supervised dimensionality reduction. In: Y. Bengio, D. Schuurmans, J. Lafferty, C. K. I. Williams and A. Culotta (Eds.): Advances in Neural Information Processing Systems 22 (NeurIPS 2009), 979-987, MIT Press, Cambridge, MA, 2010. PDF (long version)
  • A. Keller, N. Ludwig, S. Heisel, P. Leidinger, C. Andres, W.-I. Steudel, H. Huwer, B. Burgeth, M. Hein, J. Weickert, E. Meese, H.-P. Lenhof (2009): Large-scale antibody profiling of human blood sera: The future of molecular diagnosis. Informatik-Spektrum, 32:332-338. PDF
  • T. Buehler, M. Hein (2009): Spectral Clustering based on the graph p-Laplacian. In: Leon Bottou, Michael Littman (Eds.): Proceedings of the 26th International Conference on Machine Learning (ICML 2009), 81-88, Omnipress. PDF (Supplementary materialErrata to Supp. Mat.   Code
  • M. Maier, M. Hein, U. von Luxburg (2009): Optimal construction of k-nearest neighbor graphs for identifying noisy clusters. Theoretical Computer Science, 410:1749-1764. PDF

2008

  • F. Steinke, M. Hein (2008): Non-parametric Regression between Manifolds. In: D. Koller, D. Schuurmans, Y. Bengio, L. Bottou (Eds.):  Advances in Neural Information Processing Systems 21 (NeurIPS 2008), 1561 - 1568, MIT Press, Cambridge, MA, 2009. PDF
  • M. Maier, U. von Luxburg, M. Hein (2008): Influence of Graph Construction on Graph-based Clustering Measures. In: D. Koller, D. Schuurmans, Y. Bengio, L. Bottou (Eds.): Advances in Neural Information Processing Systems 21 (NeurIPS 2008), 1025 - 1032, MIT Press, Cambridge, MA, 2009. PDF
    > Markus Maier obtained for this paper the Outstanding Student Paper Award at NeurIPS 2008 <
  • M. Hein, F. Steinke, B. Schoelkopf (2008): Nonparametric regression between manifolds. Oberwolfach Report 30:34-35.
  • P. Didyk, R. Mantiuk, M. Hein, H. P. Seidel (2008): Enhancement of Bright Video Features for HDR Displays. Computer Graphics Forum, 27:1265-1274. (Proceedings of Eurographics Symposium on Rendering 2008).
  • F. Steinke, M. Hein, J. Peters, B. Schoelkopf (2008): Manifold-valued Thin-Plate Splines with Applications in Computer GraphicsComputer Graphics Forum, 27:437-448. PDF (Proceedings of EUROGRAPHICS 2008).
  • M. Hein (2008): Binary Classification under Sample Selection BiasIn: J. Quinonero Candela, M. Sugiyama, A. Schwaighofer, N. D. Lawrence (Eds.): Dataset Shift in Machine Learning. MIT Press. PDF
  • M. Hein, F. Steinke, B. Schoelkopf (2008): Energy functionals for manifold-valued mappings and their properties. Technical Report 167 (January 2008), Max Planck Institute for Biological Cybernetics. PDF

2007

  • M. Hein, J.-Y. Audibert, U. von Luxburg (2007): Graph Laplacians and their convergence on random neighborhood graphs. Journal of Machine Learning Research, 8:1325-1370. PDF
  • M. Maier, M. Hein, U. von Luxburg (2007): Cluster Identification in neighborhood graphs. In: M. Hutter, R. Servedio, E. Takimoto (Eds.): Proceedings of the 18th International Confererence on Algorithmic Learning Theory (ALT 2007), 196 - 210, Springer, New York. PDF
    > Markus Maier obtained the E. M. Gold Award (best student paper) at ALT 2007 for this paper < 
    • Corresponding technical report: 
      M. Maier, M. Hein, U. von Luxburg (2007): Cluster identification in nearest-neighbor graphs. Technical Report 163 (May 2007), Max Planck Institute for Biological Cybernetics. PDF
  • M. Hein, M. Maier (2007): Manifold Denoising for finding natural representations of data. In: Proceedings of the 22nd AAAI Conference on Artificial Intelligence (AAAI-07), 1646-1649, AAAI Press. PDF

2006

  • M. Hein, M. Maier (2006): Manifold Denoising. In: B. Schoelkopf, J. Platt,  T. Hofmann (Eds.):  Advances in Neural Information Processing Systems 19 (NeurIPS 2006), 561 - 568, MIT Press, Cambridge, MA, 2007. PDF
  • M. Hein (2006(): Uniform convergence of adaptive graph-based regularization. In: G. Lugosi,H. U. Simon (Eds.): Proceedings of the 19th Annual Conference on Learning Theory (COLT 2006), 50-64, Springer, New York. PDF

2005

  • M. Hein, O. Bousquet, B. Schoelkopf (2005): Maximal margin classification for metric spaces. Journal of Computer and System Sciences, 71:333-359. PDF
  • M. Hein, J.-Y. Audibert (2005): Intrinsic dimensionality estimation of submanifolds in Euclidean space. In: L. de Raedt, S. Wrobel (Eds.): Proceedings of the 22nd International Conference on Machine Learning (ICML 2005), 289 - 296, ACM press. PDF
  • M. Hein, J.-Y. Audibert, U. von Luxburg (2005): From graphs to manifolds - weak and strong pointwise consistency of graph Laplacians. In: R. Meir, P. Auer (Eds.): Proceedings of the 18th Conference on Learning Theory (COLT 2005), 470-485, Springer, New York. PDF
    > This paper won a best student paper award at COLT 2005 <
  • M. Hein, O. Bousquet (2005): Hilbertian metrics and positive definite kernels on probability measures. In: Z. Ghahramani, R. Cowell (Eds.): Proceedings of the 10th International Workshop on Artificial Intelligence and Statistics (AISTATS). Society for Artificial Intelligence and Statistics. PDF

2004

  • M. Hein, T. N. Lal, O. Bousquet (2004): Hilbertian metrics on probability measures and their application in SVMs. In: C. E. Rasmussen, H. H. Buelthoff, M. Giese, B. Schoelkopf (Eds.): Proceedings of the 26th DAGM Symposium. 270-277, Springer, Berlin. PDF
    • Corresponding technical report: 
      M. Hein, O. Bousquet (2004): Hilbertian metrics and positive definite kernels on probability measures. Technical Report 126 (July 2004), Max Planck Institute for Biological Cybernetics. PDF
  • M. Hein, O. Bousquet (2004): Kernels, associated structures and generalizations. Technical Report 127 (July 2004), Max Planck Institute for Biological Cybernetics. PDF

2003

  • O. Bousquet, O. Chapelle, M. Hein (2003): Measure based regularization. In: S. Thrun, L. Saul, B. Schoelkopf (Eds.): Advances in Neural Information Processing Systems 16 (NeurIPS 2003). MIT Press, Cambridge, MA, 2004. PDF
  • M. Hein, O. Bousquet (2003): Maximal margin classification for metric spaces. In: B. Schoelkopf, M. K. Warmuth (Eds.): 16th Annual Conference on Learning Theory (COLT 2003). Berlin, Springer. PDF