Search Results for author: Bálint Gyires-Tóth

Found 8 papers, 3 papers with code

Reconstructing nodal pressures in water distribution systems with graph neural networks

1 code implementation28 Apr 2021 Gergely Hajgató, Bálint Gyires-Tóth, György Paál

The reconstruction method is based on K-localized spectral graph filters, wherewith graph convolution on water networks is possible.

Deep Reinforcement Learning for Real-Time Optimization of Pumps in Water Distribution Systems

1 code implementation13 Oct 2020 Gergely Hajgató, György Paál, Bálint Gyires-Tóth

The main contribution of the presented approach is that the agent can run the pumps in real-time because it depends only on measurement data.

Predicting the flow field in a U-bend with deep neural networks

no code implementations1 Oct 2020 Gergely Hajgató, Bálint Gyires-Tóth, György Paál

This database was used to train an encoder-decoder style deep convolutional neural network to predict the velocity distribution from the geometry.

Robust Reinforcement Learning-based Autonomous Driving Agent for Simulation and Real World

no code implementations23 Sep 2020 Péter Almási, Róbert Moni, Bálint Gyires-Tóth

In our approach, the agent is trained in a simulated environment and it is able to navigate both in a simulated and real-world environment.

Autonomous Driving

Self-Attention Networks for Intent Detection

no code implementations RANLP 2019 Sevinj Yolchuyeva, Géza Németh, Bálint Gyires-Tóth

Self-attention networks (SAN) have shown promising performance in various Natural Language Processing (NLP) scenarios, especially in machine translation.

Intent Detection Machine Translation

Transformer based Grapheme-to-Phoneme Conversion

no code implementations arXiv preprint 2019 Sevinj Yolchuyeva, Géza Németh, Bálint Gyires-Tóth

The transformer network architecture is completely based on attention mechanisms, and it outperforms sequence-to-sequence models in neural machine translation without recurrent and convolutional layers.

Machine Translation Speech Recognition

Stochastic Weight Matrix-based Regularization Methods for Deep Neural Networks

2 code implementations26 Sep 2019 Patrik Reizinger, Bálint Gyires-Tóth

The aim of this paper is to introduce two widely applicable regularization methods based on the direct modification of weight matrices.

Time Series

Distance Assessment and Hypothesis Testing of High-Dimensional Samples using Variational Autoencoders

no code implementations16 Sep 2019 Marco Henrique de Almeida Inácio, Rafael Izbicki, Bálint Gyires-Tóth

Given two distinct datasets, an important question is if they have arisen from the the same data generating function or alternatively how their data generating functions diverge from one another.

Two-sample testing

Cannot find the paper you are looking for? You can Submit a new open access paper.