Search Results for author:
Found 71 papers, 37 papers with code
Face Identity-Aware Disentanglement in StyleGAN
Conditional GANs are frequently used for manipulating the attributes of face images, such as expression, hairstyle, pose, or age.
Interpretability Benchmark for Evaluating Spatial Misalignment of Prototypical Parts Explanations
Prototypical parts-based networks are becoming increasingly popular due to their faithful self-explanations.
ChiENN: Embracing Molecular Chirality with Graph Neural Networks
Graph Neural Networks (GNNs) play a fundamental role in many deep learning problems, in particular in cheminformatics.
Fantastic Weights and How to Find Them: Where to Prune in Dynamic Sparse Training
The key components of this framework are the pruning and growing criteria, which are repeatedly applied during the training process to adjust the network's sparse connectivity.
ProMIL: Probabilistic Multiple Instance Learning for Medical Imaging
The most common MIL model is when we consider a bag as positive if at least one of its instances has a positive label.
Augmentation-aware Self-supervised Learning with Guided Projector
By learning to remain invariant to applied data augmentations, methods such as SimCLR and MoCo are able to reach quality on par with supervised approaches.
MultiPlaneNeRF: Neural Radiance Field with Non-Trainable Representation
Consequently, we can only replace the 2D images (without additional training) to produce a NeRF representation of the new object.
r-softmax: Generalized Softmax with Controllable Sparsity Rate
Nowadays artificial neural network models achieve remarkable results in many disciplines.
Gaussian model for closed curves
In this work, we propose a density representation of the closed curve, which can be used to detect the complicated templates in the data.
NeRFlame: FLAME-based conditioning of NeRF for 3D face rendering
As a result, color values are non-zero only in the proximity of the FLAME mesh.
Contrastive Hierarchical Clustering
Deep clustering has been dominated by flat models, which split a dataset into a predefined number of groups.
Ranked #1 on
Image Clustering
on MNIST
Hypernetworks build Implicit Neural Representations of Sounds
Implicit Neural Representations (INRs) are nowadays used to represent multimedia signals across various real-life applications, including image super-resolution, image compression, or 3D rendering.
Revisiting Offline Compression: Going Beyond Factorization-based Methods for Transformer Language Models
Recent transformer language models achieve outstanding results in many natural language processing (NLP) tasks.
ProtoSeg: Interpretable Semantic Segmentation with Prototypical Parts
We introduce ProtoSeg, a novel model for interpretable semantic image segmentation, which constructs its predictions using similar patches from the training set.
HyperSound: Generating Implicit Neural Representations of Audio Signals with Hypernetworks
Implicit neural representations (INRs) are a rapidly growing research field, which provides alternative ways to represent multimedia signals.
Hypernetwork approach to Bayesian MAML
Unfortunately, previous modifications of MAML are limited due to the simplicity of Gaussian posteriors, MAML-like gradient-based weight updates, or by the same structure enforced for universal and adapted weights.
ProPaLL: Probabilistic Partial Label Learning
Partial label learning is a type of weakly supervised learning, where each training instance corresponds to a set of candidate labels, among which only one is true.
LIDL: Local Intrinsic Dimension Estimation Using Approximate Likelihood
We attempt to address that challenge by proposing a novel approach to the problem: Local Intrinsic Dimension estimation using approximate Likelihood (LIDL).
SLOVA: Uncertainty Estimation Using Single Label One-Vs-All Classifier
The proposed classifier is confident only if a single class has a high probability and other probabilities are negligible.
Bounding Evidence and Estimating Log-Likelihood in VAE
Many crucial problems in deep learning and statistics are caused by a variational gap, i. e., a difference between evidence and evidence lower bound (ELBO).
Continual Learning with Guarantees via Weight Interval Constraints
We introduce a new training paradigm that enforces interval constraints on neural network parameter space to control forgetting.
HyperShot: Few-Shot Learning by Kernel HyperNetworks
Few-shot models aim at making predictions using a minimal number of labeled examples from a given task.
Interpretable Image Classification with Differentiable Prototypes Assignment
We introduce ProtoPool, an interpretable image classification model with a pool of prototypes shared by the classes.
MisConv: Convolutional Neural Networks for Missing Data
Processing of missing data by modern neural networks, such as CNNs, remains a fundamental, yet unsolved challenge, which naturally arises in many practical applications, like image inpainting or autonomous vehicles and robots.
Non-Gaussian Gaussian Processes for Few-Shot Regression
This makes the GP posterior locally non-Gaussian, therefore we name our method Non-Gaussian Gaussian Processes (NGGPs).
Relative Molecule Self-Attention Transformer
Self-supervised learning holds promise to revolutionize molecule property prediction - a central task to drug discovery and many more industries - by enabling data efficient learning from scarce experimental data.
Flow-based SVDD for anomaly detection
We propose FlowSVDD -- a flow-based one-class classifier for anomaly/outliers detection that realizes a well-known SVDD principle using deep learning tools.
SONG: Self-Organizing Neural Graphs
Recent years have seen a surge in research on deep interpretable neural networks with decision trees as one of the most commonly incorporated tools.
Direction is what you need: Improving Word Embedding Compression in Large Language Models
The adoption of Transformer-based models in natural language processing (NLP) has led to great success using a massive number of parameters.
Zero Time Waste: Recycling Predictions in Early Exit Neural Networks
The problem of reducing processing time of large deep learning models is a fundamental challenge in many real-world applications.
HyperPocket: Generative Point Cloud Completion
In this work, we reformulate the problem of point cloud completion into an object hallucination task.
RegFlow: Probabilistic Flow-based Regression for Future Prediction
Predicting future states or actions of a given system remains a fundamental, yet unsolved challenge of intelligence, especially in the scope of complex and non-deterministic scenarios, such as modeling behavior of humans.
ProtoPShare: Prototype Sharing for Interpretable Image Classification and Similarity Discovery
In this paper, we introduce ProtoPShare, a self-explained method that incorporates the paradigm of prototypical parts to explain its predictions.
OneFlow: One-class flow for anomaly detection based on a minimal volume region
We propose OneFlow - a flow-based one-class classifier for anomaly (outlier) detection that finds a minimal volume bounding region.
Generative models with kernel distance in data space
First, an autoencoder based architecture, using kernel measures, is built to model a manifold of data.
Adversarial Examples Detection and Analysis with Layer-wise Autoencoders
We present a mechanism for detecting adversarial examples based on data representations taken from the hidden layers of the target network.
HyperFlow: Representing 3D Objects as Surfaces
To that end, we devise a generative model that uses a hypernetwork to return the weights of a Continuous Normalizing Flows (CNF) target network.
Kernel Self-Attention in Deep Multiple Instance Learning
There have been several attempts to create a model working with a bag of instances, however, they are assuming that there are no dependencies within the bag and the label is connected to at least one instance.
Finding the Optimal Network Depth in Classification Tasks
We develop a fast end-to-end method for training lightweight neural networks using multiple classifier heads.
The Break-Even Point on Optimization Trajectories of Deep Neural Networks
We argue for the existence of the "break-even" point on this trajectory, beyond which the curvature of the loss surface and noise in the gradient are implicitly regularized by SGD.
Molecule Attention Transformer
Designing a single neural network architecture that performs competitively across a range of molecule property prediction tasks remains largely an open challenge, and its solution may unlock a widespread use of deep learning in the drug discovery industry.
LocoGAN -- Locally Convolutional GAN
In the paper we construct a fully convolutional GAN model: LocoGAN, which latent space is given by noise-like images of possibly different resolutions.
Hypernetwork approach to generating point clouds
The main idea of our HyperCloud method is to build a hyper network that returns weights of a particular neural network (target network) trained to map points from a uniform unit ball distribution into a 3D shape.
WICA: nonlinear weighted ICA
Independent Component Analysis (ICA) aims to find a coordinate system in which the components of the data are independent.
Biologically-Inspired Spatial Neural Networks
We introduce bio-inspired artificial neural networks consisting of neurons that are additionally characterized by spatial positions.
Spatial Graph Convolutional Networks
Graph Convolutional Networks (GCNs) have recently become the primary choice for learning from graph-structured data, superseding hash fingerprints in representing chemical compounds.
SeGMA: Semi-Supervised Gaussian Mixture Auto-Encoder
We propose a semi-supervised generative model, SeGMA, which learns a joint probability distribution of data and their classes and which is implemented in a typical Wasserstein auto-encoder framework.
Fast and Stable Interval Bounds Propagation for Training Verifiably Robust Models
We present an efficient technique, which allows to train classification networks which are verifiably robust against norm-bounded adversarial attacks.
Independent Component Analysis based on multiple data-weighting
Independent Component Analysis (ICA) - one of the basic tools in data analysis - aims to find a coordinate system in which the components of the data are independent.
Feature-Based Interpolation and Geodesics in the Latent Spaces of Generative Models
In the case of geodesics, we search for the curves with the shortest length, while in the case of generative models we typically apply linear interpolation in the latent space.
Non-linear ICA based on Cramer-Wold metric
Non-linear source separation is a challenging open problem with many applications.
Hypernetwork functional image representation
Motivated by the human way of memorizing images we introduce their functional representation, where an image is represented by a neural network.
LOSSGRAD: automatic learning rate in gradient descent
In this paper, we propose a simple, fast and easy to implement algorithm LOSSGRAD (locally optimal step-size in gradient descent), which automatically modifies the step-size in gradient descent during neural networks training.
Sliced generative models
In this paper we discuss a class of AutoEncoder based generative models based on one dimensional sliced approach.
Set Aggregation Network as a Trainable Pooling Layer
Global pooling, such as max- or sum-pooling, is one of the key ingredients in deep neural networks used for processing images, texts, graphs and other types of structured data.
Deep processing of structured data
We construct a general unified framework for learning representation of structured data, i. e. data which cannot be represented as the fixed-length vectors (e. g. sets, graphs, texts or images of varying sizes).
Dynamical Isometry is Achieved in Residual Networks in a Universal Way for any Activation Function
We propose that in ResNets this can be resolved based on our results, by ensuring the same level of dynamical isometry at initialization.
Cramer-Wold AutoEncoder
The crucial new ingredient is the introduction of a new (Cramer-Wold) metric in the space of densities, which replaces the Wasserstein metric used in SWAE.
Processing of missing data by neural networks
We propose a general, theoretically justified mechanism for processing missing data by neural networks.
Cascade context encoder for improved inpainting
For this purpose, we train context encoder for 64x64 pixels images in a standard way and use its resized output to fill in the missing input region of the 128x128 context encoder, both in training and evaluation phase.
Efficient mixture model for clustering of sparse high dimensional binary data
In this paper we propose a mixture model, SparseMix, for clustering of sparse high dimensional binary data, which connects model-based with centroid-based clustering.
Semi-supervised model-based clustering with controlled clusters leakage
In this paper, we focus on finding clusters in partially categorized data sets.
Spherical Wards clustering and generalized Voronoi diagrams
In order to graphically represent and interpret the results the notion of Voronoi diagram was generalized to non Euclidean spaces and applied for introduced clustering method.
Pointed subspace approach to incomplete data
Incomplete data are often represented as vectors with filled missing attributes joined with flag vectors indicating missing components.
Generalized RBF kernel for incomplete data
We construct $\bf genRBF$ kernel, which generalizes the classical Gaussian RBF kernel to the case of incomplete data.
Introduction to Cross-Entropy Clustering The R Package CEC
The R Package CEC performs clustering based on the cross-entropy clustering (CEC) method, which was recently developed with the use of information theory.
Maximum Entropy Linear Manifold for Learning Discriminative Low-dimensional Representation
Representation learning is currently a very hot topic in modern machine learning, mostly due to the great success of the deep learning methods.
Extreme Entropy Machines: Robust information theoretic classification
The main contribution of this paper is proposing a model based on the information theoretic concepts which on the one hand shows new, entropic perspective on known linear classifiers and on the other leads to a construction of very robust method competetitive with the state of the art non-information theoretic ones (including Support Vector Machines and Extreme Learning Machines).
Cluster based RBF Kernel for Support Vector Machines
In the classical Gaussian SVM classification we use the feature space projection transforming points to normal distributions with fixed covariance matrices (identity in the standard RBF and the covariance of the whole dataset in Mahalanobis RBF).
Multithreshold Entropy Linear Classifier
Then we prove that our method is a multithreshold large margin classifier, which shows the analogy to the SVM, while in the same time works with much broader class of hypotheses.
