Michele Scarpiniti

Pubblicazioni

Titolo Pubblicato in Anno
Deep recurrent neural networks for audio classification in construction sites European signal processing conference 2021
A wide multimodal dense U-net for fast magnetic resonance imaging European signal processing conference 2021
MATLAB® per l'audio 2021
Learning-in-the-Fog (LiFo): Deep learning meets Fog Computing for the minimum-energy distributed early-exit of inference in delay-critical IoT realms IEEE ACCESS 2021
Deep Belief Network based audio classification for construction sites monitoring EXPERT SYSTEMS WITH APPLICATIONS 2021
Introduzione all'audio real-time: Basi teoriche e prime applicazioni 2021
A histogram-based low-complexity approach for the effective detection of COVID-19 disease from CT and X-ray images APPLIED SCIENCES 2021
Music genre classification using stacked auto-encoders Smart Innovation, Systems and Technologies 2020
Optimized training and scalable implementation of Conditional Deep Neural Networks with early exits for Fog-supported IoT applications INFORMATION SCIENCES 2020
Advanced sound classifiers and performance analyses for accurate audio-based construction project monitoring JOURNAL OF COMPUTING IN CIVIL ENGINEERING 2020
A CNN approach for audio classification in construction sites Progresses in Artificial Intelligence and Neural Systems 2020
Differentiable branching in deep networks for fast inference ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings 2020
Why should we add early exits to neural networks? COGNITIVE COMPUTATION 2020
Laboratorio di programmazione. MATLAB® per l'ingegnere dell'informazione 2020
Steady-state performance of an adaptive combined MISO filter using the multichannel affine projection algorithm ALGORITHMS 2019
Separation of drum and bass from monaural tracks Smart Innovation, Systems and Technologies 2019
On 4-dimensional hypercomplex algebras in adaptive signal processing Smart Innovation, Systems and Technologies 2019
Learning activation functions from data using cubic spline interpolation Smart Innovation, Systems and Technologies 2019
A low-complexity linear-in-the-parameters nonlinear filter for distorted speech signals Smart Innovation, Systems and Technologies 2019
VirtFogSim: A parallel toolbox for dynamic energy-delay performance testing and optimization of 5G Mobile-Fog-Cloud virtualized platforms APPLIED SCIENCES 2019

ERC

  • PE6_11
  • PE7_7

KET

  • Big data & computing

Interessi di ricerca

Gli attuali interessi di ricerca sono nel campo dell'elaborazione non lineare del segnale, dei circuiti e degli algoritmi per l'elaborazione del segnale audio, array processing e la separazione di sorgenti. Inoltre, è attivo su tematiche di filtraggio adattativo non lineare, con particolare enfasi all'identificazione di sistemi non lineari. Altri interessi di ricerca riguardano la cancellazione dell'eco monofonico/stereofonico in ambiente avverso e in presenza di forti distorsioni non lineari e la localizzazione acustica di sorgenti all’interno di ambienti riverberanti. Si interessa anche di machine learning e di reti neurali per l'elaborazione del segnale.

Keywords

audio
adaptive signal processing
complex nonlinear filters
energy aware machine learning
fog computing (FC)

Progetti di Ricerca

© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma