dense lattice
Recently Published Documents


TOTAL DOCUMENTS

17
(FIVE YEARS 4)

H-INDEX

4
(FIVE YEARS 1)

2020 ◽  
Vol 34 (05) ◽  
pp. 8552-8559
Author(s):  
Mohammad Nikzad ◽  
Aaron Nicolson ◽  
Yongsheng Gao ◽  
Jun Zhou ◽  
Kuldip K. Paliwal ◽  
...  

Convolutional neural networks (CNNs) with residual links (ResNets) and causal dilated convolutional units have been the network of choice for deep learning approaches to speech enhancement. While residual links improve gradient flow during training, feature diminution of shallow layer outputs can occur due to repetitive summations with deeper layer outputs. One strategy to improve feature re-usage is to fuse both ResNets and densely connected CNNs (DenseNets). DenseNets, however, over-allocate parameters for feature re-usage. Motivated by this, we propose the residual-dense lattice network (RDL-Net), which is a new CNN for speech enhancement that employs both residual and dense aggregations without over-allocating parameters for feature re-usage. This is managed through the topology of the RDL blocks, which limit the number of outputs used for dense aggregations. Our extensive experimental investigation shows that RDL-Nets are able to achieve a higher speech enhancement performance than CNNs that employ residual and/or dense aggregations. RDL-Nets also use substantially fewer parameters and have a lower computational requirement. Furthermore, we demonstrate that RDL-Nets outperform many state-of-the-art deep learning approaches to speech enhancement. Availability: https://github.com/nick-nikzad/RDL-SE.


2019 ◽  
Vol 55 (11) ◽  
Author(s):  
Alexei Bazavov ◽  
◽  
Frithjof Karsch ◽  
Swagato Mukherjee ◽  
Peter Petreczky
Keyword(s):  

2016 ◽  
Vol 109 (4) ◽  
pp. 042406 ◽  
Author(s):  
M. V. Sapozhnikov ◽  
S. N. Vdovichev ◽  
O. L. Ermolaeva ◽  
N. S. Gusev ◽  
A. A. Fraerman ◽  
...  
Keyword(s):  

2010 ◽  
Vol 638-642 ◽  
pp. 1143-1147 ◽  
Author(s):  
Koji Tanaka ◽  
Tetsu Kiyobayashi ◽  
Nobuhiko Takeichi ◽  
Hiroshi Miyamura ◽  
Shiomi Kikuchi

Super-laminates have been attracting attention since co-authors Ueda et al. reported that Mg/Cu super-laminates showed reversible hydrogenation and dehydrogenation at 473K. The Mg/Cu super-laminates were prepared by a repetitive fold and roll method. Initial activation at 573 K led the super-laminates to absorb hydrogen at 473K. TEM observations of micro/nano-structures in the Mg/Cu super-laminates and Mg2Cu powder were performed in order to clarify the process of hydrogenation and dehydrogenation at 473K. The as-rolled Mg/Cu super-laminates have laminated structures sub-micrometer thick composed of Mg and Cu layers with dense lattice defects. The super-laminates after initial activation keep laminated structure and have uniformly distributed pores sub-micrometer in diameter. It is considered that these micro/nano-structures of Mg/Cu super-laminates lead to lower dehydrogenation temperature and better kinetics.


2009 ◽  
Vol 2 (1) ◽  
pp. 87-90
Author(s):  
R. M. H. Rahman

By a new partial ordering relation "≤" the set of convex sublattices CS(L)  of a lattice L is again a lattice. In this paper we establish some results on the pseudocomplementation of (CS(L); ≤). We show that a lattice L with 0 is dense if and only if CS(L) is dense. Then we prove that a finite distributive lattice is a Stone lattice if and only if CS(L) is Stone. We also prove that an upper continuous lattice L is a Stone lattice if and only if CS(L) is Stone.  Keywords: Upper continuous lattice; Pseudocomplemented lattice; Dense lattice; Stone lattice. © 2010 JSR Publications. ISSN: 2070-0237 (Print); 2070-0245 (Online). All rights reserved.  DOI: 10.3329/jsr.v2i1.2485                  J. Sci. Res. 2 (1), 87-90  (2010) 


2007 ◽  
Vol 561-565 ◽  
pp. 1581-1584
Author(s):  
Koji Tanaka ◽  
Nobuhiko Takeichi ◽  
Hideaki Tanaka ◽  
Nobuhiro Kuriyama ◽  
Tamotsu T. Ueda ◽  
...  

Super-laminates have been attracting attention since co-authors Ueda et al. reported that Mg/Cu super-laminates showed reversible hydrogenation and dehydrogenation at 473K. The Mg/Cu super-laminates were prepared by a repetitive fold and roll method. Initial activation at 573 K led the super-laminates to absorb hydrogen at 473K. TEM observations of micro/nano-structures in the super-laminates and Mg2Cu powder were performed in order to clarify the process of hydrogenation and dehydrogenation at 473K. The as-rolled Mg/Cu super-laminates have laminated structures in size of sub-micrometer thickness composed of Mg and Cu layers with dense lattice defects. The super-laminates after initial activation keep laminated structure and have uniformly distributed pores with a sub-micrometer diameter. It is considered that these micro/nano-structures of Mg/Cu super-laminates lead to lower dehydrogenation temperature and better kinetics.


Sign in / Sign up

Export Citation Format

Share Document