EUSIPCO 2020

120 papers
The 2020 edition of the European Signal Processing Conference (EUSIPCO) featured 127 papers focusing on advancements in decentralized optimization, signal recovery, and neural network efficiency. A primary research trend involved robust optimization in dynamic network topologies and privacy-preserving distributed graph filtering. Audio processing remained a significant theme, with studies exploring asymmetric and dilated convolutions for time-frequency modeling, as well as Bayesian hierarchical models for blind source separation. Additional research addressed sparse signal recovery using heavy-tailed priors and statistical inference via characteristic functions. The conference also featured work on image reconstruction techniques that move beyond pixel-wise loss and methods for separating physiological signals with spectral overlap. Notable contributions came from authors such as Alejandro Ribeiro, Nikos Deligiannis, and Geert Leus. Specific applications discussed included monarch butterfly localization via multi-modal sensor fusion, data hiding in JPEG crypto-compressed images, and audio classification for construction sites using deep recurrent neural networks. These works reflect a broader shift toward integrating robust optimization with complex signal modeling across diverse physical and social domains.

Top Research Directions

Research directions most represented by EUSIPCO 2020 papers.

Learning Latent Graph Topologies Directly from Data 2
Deep Unfolding Networks for Inverse Problems 2
Parameter Linearization for Least Squares Estimation 1
Learning Dynamical Systems via Markov Parameter Realization 1
Topology Inference via Graph Signal Stationarity 1
The Mechanics of Stubbornness in Social Networks 1
Adaptive Dictionary Maintenance for Online Kernel Learning 1
Conjugate Bayesian Inference via Pólya Gamma Augmentation 1
Overcoming Noise Overfitting in Untrained Deep Priors 1
Continuous Graph Limits for Transferable Neural Networks 1
Hardware and Computational Bottlenecks of Complex Activations 1
Non-Uniform Frequency Processing for Efficient Neural Models 1
Nonlinear Spectral Unmixing for Hyperspectral Imaging 1
Image Reconstruction via Coupled Dictionary Learning 1
Breaking Translation Invariance in Spectrogram Convolutions 1

Top Authors

AuthorPapers
Felix Krahmer4
Sharon Gannot4
Alejandro Ribeiro3
Elvin Isufi3
Nikos Deligiannis3
Aggelos Katsaggelos2
Arash Mohammadi2
Aurelio Uncini2
Axel Roebel2
Danilo Comminiello2
Konstantinos N. Plataniotis2
Luana Ruiz2
Maitreya Patel2
Martin B. Møller2
Mike Davies2
Mingyu Yang2
Nada Sissouno2
Nicolas Gillis2
Pablo Martínez-Nuevo2
Qiongxiu Li2

Impactful Papers

Papers from EUSIPCO 2020 that shaped research clusters.

Foundational

Influential

Papers

Title
Semi-supervised learning of glottal pulse positions in a neural analysis-synthesis framework
Semi-supervised Multichannel Speech Separation Based on a Phone- and Speaker-Aware Deep Generative Model of Speech Spectrograms
Signal-Adapted Analytic Wavelet Packets in Arbitrary Dimensions
Sound Event Localization and Detection Using Convolutional Recurrent Neural Networks and Gated Linear Units
Speech Dereverberation Performance of a Polynomial-EVD Subspace Approach
Stable Ptychographic Phase Retrieval via Lost Subspace Completion
State-Space Based Network Topology Identification
Stochasticity and Skip Connection Improve Knowledge Transfer
Successive Nonnegative Projection Algorithm for Linear Quadratic Mixtures
Super-Resolution Time-of-Arrival Estimation using Neural Networks
Taco-VC: A Single Speaker Tacotron based Voice Conversion with Limited Data
Tensor Recovery via Nonconvex Low-Rank Approximation
The Benefits of Side Information for Structured Phase Retrieval
The Modulo Radon Transform and its Inversion
Time Encoding Using the Hyperbolic Secant Kernel
Towards Finite-Time Consensus with Graph Convolutional Neural Networks
Type/position classification of inter-floor noise in residential buildings with a single microphone via supervised learning
Wavelets in the Deep Learning Era
WaveNet based architectures for denoising periodic discontinuous signals and application to friction signals
Weak Speech Supervision: A case study of Dysarthria Severity Classification

« prev Page 3/3