Alert button
Picture for Aleksandr Petiushko

Aleksandr Petiushko

Alert button

NUQ: Nonparametric Uncertainty Quantification for Deterministic Neural Networks

Feb 07, 2022
Nikita Kotelevskii, Aleksandr Artemenkov, Kirill Fedyanin, Fedor Noskov, Alexander Fishkov, Aleksandr Petiushko, Maxim Panov

Figure 1 for NUQ: Nonparametric Uncertainty Quantification for Deterministic Neural Networks
Figure 2 for NUQ: Nonparametric Uncertainty Quantification for Deterministic Neural Networks
Figure 3 for NUQ: Nonparametric Uncertainty Quantification for Deterministic Neural Networks
Figure 4 for NUQ: Nonparametric Uncertainty Quantification for Deterministic Neural Networks

This paper proposes a fast and scalable method for uncertainty quantification of machine learning models' predictions. First, we show the principled way to measure the uncertainty of predictions for a classifier based on Nadaraya-Watson's nonparametric estimate of the conditional label distribution. Importantly, the approach allows to disentangle explicitly aleatoric and epistemic uncertainties. The resulting method works directly in the feature space. However, one can apply it to any neural network by considering an embedding of the data induced by the network. We demonstrate the strong performance of the method in uncertainty estimation tasks on a variety of real-world image datasets, such as MNIST, SVHN, CIFAR-100 and several versions of ImageNet.

Viaarxiv icon

Smoothed Embeddings for Certified Few-Shot Learning

Feb 02, 2022
Mikhail Pautov, Olesya Kuznetsova, Nurislam Tursynbek, Aleksandr Petiushko, Ivan Oseledets

Figure 1 for Smoothed Embeddings for Certified Few-Shot Learning
Figure 2 for Smoothed Embeddings for Certified Few-Shot Learning
Figure 3 for Smoothed Embeddings for Certified Few-Shot Learning
Figure 4 for Smoothed Embeddings for Certified Few-Shot Learning

Randomized smoothing is considered to be the state-of-the-art provable defense against adversarial perturbations. However, it heavily exploits the fact that classifiers map input objects to class probabilities and do not focus on the ones that learn a metric space in which classification is performed by computing distances to embeddings of classes prototypes. In this work, we extend randomized smoothing to few-shot learning models that map inputs to normalized embeddings. We provide analysis of Lipschitz continuity of such models and derive robustness certificate against $\ell_2$-bounded perturbations that may be useful in few-shot learning scenarios. Our theoretical results are confirmed by experiments on different datasets.

Viaarxiv icon

Many Heads but One Brain: an Overview of Fusion Brain Challenge on AI Journey 2021

Nov 22, 2021
Daria Bakshandaeva, Denis Dimitrov, Alex Shonenkov, Mark Potanin, Vladimir Arkhipkin, Denis Karachev, Vera Davydova, Anton Voronov, Mikhail Martynov, Natalia Semenova, Mikhail Stepnov, Elena Tutubalina, Andrey Chertok, Aleksandr Petiushko

Figure 1 for Many Heads but One Brain: an Overview of Fusion Brain Challenge on AI Journey 2021
Figure 2 for Many Heads but One Brain: an Overview of Fusion Brain Challenge on AI Journey 2021
Figure 3 for Many Heads but One Brain: an Overview of Fusion Brain Challenge on AI Journey 2021
Figure 4 for Many Heads but One Brain: an Overview of Fusion Brain Challenge on AI Journey 2021

Supporting the current trend in the AI community, we propose the AI Journey 2021 Challenge called Fusion Brain which is targeted to make the universal architecture process different modalities (namely, images, texts, and code) and to solve multiple tasks for vision and language. The Fusion Brain Challenge https://github.com/sberbank-ai/fusion_brain_aij2021 combines the following specific tasks: Code2code Translation, Handwritten Text recognition, Zero-shot Object Detection, and Visual Question Answering. We have created datasets for each task to test the participants' submissions on it. Moreover, we have opened a new handwritten dataset in both Russian and English, which consists of 94,130 pairs of images and texts. The Russian part of the dataset is the largest Russian handwritten dataset in the world. We also propose the baseline solution and corresponding task-specific solutions as well as overall metrics.

Viaarxiv icon

CC-Cert: A Probabilistic Approach to Certify General Robustness of Neural Networks

Sep 22, 2021
Mikhail Pautov, Nurislam Tursynbek, Marina Munkhoeva, Nikita Muravev, Aleksandr Petiushko, Ivan Oseledets

Figure 1 for CC-Cert: A Probabilistic Approach to Certify General Robustness of Neural Networks
Figure 2 for CC-Cert: A Probabilistic Approach to Certify General Robustness of Neural Networks
Figure 3 for CC-Cert: A Probabilistic Approach to Certify General Robustness of Neural Networks
Figure 4 for CC-Cert: A Probabilistic Approach to Certify General Robustness of Neural Networks

In safety-critical machine learning applications, it is crucial to defend models against adversarial attacks -- small modifications of the input that change the predictions. Besides rigorously studied $\ell_p$-bounded additive perturbations, recently proposed semantic perturbations (e.g. rotation, translation) raise a serious concern on deploying ML systems in real-world. Therefore, it is important to provide provable guarantees for deep learning models against semantically meaningful input transformations. In this paper, we propose a new universal probabilistic certification approach based on Chernoff-Cramer bounds that can be used in general attack settings. We estimate the probability of a model to fail if the attack is sampled from a certain distribution. Our theoretical findings are supported by experimental results on different datasets.

Viaarxiv icon

Manifold Hypothesis in Data Analysis: Double Geometrically-Probabilistic Approach to Manifold Dimension Estimation

Jul 08, 2021
Alexander Ivanov, Gleb Nosovskiy, Alexey Chekunov, Denis Fedoseev, Vladislav Kibkalo, Mikhail Nikulin, Fedor Popelenskiy, Stepan Komkov, Ivan Mazurenko, Aleksandr Petiushko

Figure 1 for Manifold Hypothesis in Data Analysis: Double Geometrically-Probabilistic Approach to Manifold Dimension Estimation
Figure 2 for Manifold Hypothesis in Data Analysis: Double Geometrically-Probabilistic Approach to Manifold Dimension Estimation
Figure 3 for Manifold Hypothesis in Data Analysis: Double Geometrically-Probabilistic Approach to Manifold Dimension Estimation
Figure 4 for Manifold Hypothesis in Data Analysis: Double Geometrically-Probabilistic Approach to Manifold Dimension Estimation

Manifold hypothesis states that data points in high-dimensional space actually lie in close vicinity of a manifold of much lower dimension. In many cases this hypothesis was empirically verified and used to enhance unsupervised and semi-supervised learning. Here we present new approach to manifold hypothesis checking and underlying manifold dimension estimation. In order to do it we use two very different methods simultaneously - one geometric, another probabilistic - and check whether they give the same result. Our geometrical method is a modification for sparse data of a well-known box-counting algorithm for Minkowski dimension calculation. The probabilistic method is new. Although it exploits standard nearest neighborhood distance, it is different from methods which were previously used in such situations. This method is robust, fast and includes special preliminary data transformation. Experiments on real datasets show that the suggested approach based on two methods combination is powerful and effective.

Viaarxiv icon

Darker than Black-Box: Face Reconstruction from Similarity Queries

Jul 02, 2021
Anton Razzhigaev, Klim Kireev, Igor Udovichenko, Aleksandr Petiushko

Figure 1 for Darker than Black-Box: Face Reconstruction from Similarity Queries
Figure 2 for Darker than Black-Box: Face Reconstruction from Similarity Queries
Figure 3 for Darker than Black-Box: Face Reconstruction from Similarity Queries
Figure 4 for Darker than Black-Box: Face Reconstruction from Similarity Queries

Several methods for inversion of face recognition models were recently presented, attempting to reconstruct a face from deep templates. Although some of these approaches work in a black-box setup using only face embeddings, usually, on the end-user side, only similarity scores are provided. Therefore, these algorithms are inapplicable in such scenarios. We propose a novel approach that allows reconstructing the face querying only similarity scores of the black-box model. While our algorithm operates in a more general setup, experiments show that it is query efficient and outperforms the existing methods.

Viaarxiv icon

Certified Robustness via Randomized Smoothing over Multiplicative Parameters

Jun 28, 2021
Nikita Muravev, Aleksandr Petiushko

Figure 1 for Certified Robustness via Randomized Smoothing over Multiplicative Parameters
Figure 2 for Certified Robustness via Randomized Smoothing over Multiplicative Parameters
Figure 3 for Certified Robustness via Randomized Smoothing over Multiplicative Parameters
Figure 4 for Certified Robustness via Randomized Smoothing over Multiplicative Parameters

We propose a novel approach of randomized smoothing over multiplicative parameters. Using this method we construct certifiably robust classifiers with respect to a gamma-correction perturbation and compare the result with classifiers obtained via Gaussian smoothing. To the best of our knowledge it is the first work concerning certified robustness against the multiplicative gamma-correction transformation.

Viaarxiv icon

MDMMT: Multidomain Multimodal Transformer for Video Retrieval

Mar 19, 2021
Maksim Dzabraev, Maksim Kalashnikov, Stepan Komkov, Aleksandr Petiushko

Figure 1 for MDMMT: Multidomain Multimodal Transformer for Video Retrieval
Figure 2 for MDMMT: Multidomain Multimodal Transformer for Video Retrieval
Figure 3 for MDMMT: Multidomain Multimodal Transformer for Video Retrieval
Figure 4 for MDMMT: Multidomain Multimodal Transformer for Video Retrieval

We present a new state-of-the-art on the text to video retrieval task on MSRVTT and LSMDC benchmarks where our model outperforms all previous solutions by a large margin. Moreover, state-of-the-art results are achieved with a single model on two datasets without finetuning. This multidomain generalisation is achieved by a proper combination of different video caption datasets. We show that training on different datasets can improve test results of each other. Additionally we check intersection between many popular datasets and found that MSRVTT has a significant overlap between the test and the train parts, and the same situation is observed for ActivityNet.

Viaarxiv icon

Quadric hypersurface intersection for manifold learning in feature space

Feb 11, 2021
Fedor Pavutnitskiy, Sergei O. Ivanov, Evgeny Abramov, Viacheslav Borovitskiy, Artem Klochkov, Viktor Vialov, Anatolii Zaikovskii, Aleksandr Petiushko

Figure 1 for Quadric hypersurface intersection for manifold learning in feature space
Figure 2 for Quadric hypersurface intersection for manifold learning in feature space
Figure 3 for Quadric hypersurface intersection for manifold learning in feature space
Figure 4 for Quadric hypersurface intersection for manifold learning in feature space

The knowledge that data lies close to a particular submanifold of the ambient Euclidean space may be useful in a number of ways. For instance, one may want to automatically mark any point far away from the submanifold as an outlier, or to use its geodesic distance to measure similarity between points. Classical problems for manifold learning are often posed in a very high dimension, e.g. for spaces of images or spaces of representations of words. Today, with deep representation learning on the rise in areas such as computer vision and natural language processing, many problems of this kind may be transformed into problems of moderately high dimension, typically of the order of hundreds. Motivated by this, we propose a manifold learning technique suitable for moderately high dimension and large datasets. The manifold is learned from the training data in the form of an intersection of quadric hypersurfaces -- simple but expressive objects. At test time, this manifold can be used to introduce an outlier score for arbitrary new points and to improve a given similarity metric by incorporating learned geometric structure into it.

Viaarxiv icon