2 Hits in 1.2 sec

MisConv: Convolutional Neural Networks for Missing Data [article]

Marcin Przewięźlikowski, Marek Śmieja, Łukasz Struski, Jacek Tabor
2021 arXiv   pre-print
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  ...  By modeling the distribution of missing values by the Mixture of Factor Analyzers, we cover the spectrum of possible replacements and find an analytical formula for the expected value of convolution operator  ...  For the purpose of Open Access, the authors have applied a CC-BY public copyright licence to any Author Accepted Manuscript (AAM) version arising from this submission.  ... 
arXiv:2110.14010v2 fatcat:5l6kll3qevhyzdi4oklrnkfjmu

Cross-relation based blind identification of acoustic SIMO systems and applications

Mathieu Hu, Patrick A. Naylor, Mike Brookes, European Union
The increased use of devices controlled by distant speech therefore induces the need for dereverberation.  ...  From a more practical perspective, the use of room impulses estimated at a poor accuracy is investigated for the problem of speaker diarization. The spatial information c [...]  ...  However, according to results found in the literature of neural networks, the NMCFLMS is guaranteed to converge in noisy conditions only when a non-summable step-size decreasing to 0 is employed, which  ... 
doi:10.25560/52430 fatcat:ie3nbr3lqnbltjhfgeu5arku6y