Research Explorer

Uw browser ondersteunt geen JavaScript of JavaScript is niet ingeschakeld. Zonder JavaScript kan sommige functionaliteit van deze webapplicatie uitgeschakeld zijn of foutmeldingen veroorzaken. Raadpleeg om JavaScript in te schakelen de handleiding van uw browser of contacteer uw systeembeheerder.

Project

Faciliteren van communicatie onder ongunstige omstandigheden door de exploitatie van geschatte directions-of-arrival en andere a-priori kennis voor de gezamenlijke scheiding van akoestische bronnen en dereverberatie

Output en outcomes

Publicaties

Weakly DOA guided speaker separation with random look directions and iteratively refined target and interference priors

Alexander BohlenderAnn SprietWouter TirryNilesh Madhu

Insights into magnitude and phase estimation by masking and mapping in DNN-based multichannel speaker separation

Alexander BohlenderAnn SprietWouter TirryNilesh Madhu

Spatially selective speaker separation using a DNN with a location dependent feature extraction

Alexander BohlenderAnn SprietWouter TirryNilesh Madhu

A1 Artikel in een tijdschrift in IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING

CRNN-based multi-DOA estimator : comparing classification and regression

Pieter CooremanAlexander BohlenderNilesh Madhu

Improved deep speaker localization and tracking : revised training paradigm and controlled latency

Alexander BohlenderLiesbeth RoelensNilesh Madhu

Improved separation of closely-spaced speakers by exploiting auxiliary direction of arrival information within a U-net architecture

Stijn KindtAlexander BohlenderNilesh Madhu

DOA-guided source separation with direction-based initialization and time annotations using complex angular central Gaussian mixture models

Alexander BohlenderLucas Van SeverenJonathan SterckxNilesh Madhu

A1 Artikel in een tijdschrift in EURASIP JOURNAL ON AUDIO SPEECH AND MUSIC PROCESSING

2D acoustic source localisation using decentralised deep neural networks on distributed microphone arrays

Stijn KindtAlexander BohlenderNilesh Madhu

Neural networks using full-band and subband spatial features for mask based source separation

Alexander BohlenderAnn SprietWouter TirryNilesh Madhu