scispace - formally typeset
Search or ask a question
Institution

Helsinki University of Technology

About: Helsinki University of Technology is a based out in . It is known for research contribution in the topics: Thin film & Vortex. The organization has 8962 authors who have published 20136 publications receiving 723787 citations. The organization is also known as: TKK & Teknillinen korkeakoulu.


Papers
More filters
Journal ArticleDOI
TL;DR: In this paper, the functional counterpoise method was applied to the theoretical prediction of hydrogen bonding potential surfaces, using a minimal basis to represent the atomic orbitals (STO-3G).
Abstract: The “functional” counterpoise method, proposed by Boys and Bernardi [1], is applied to the theoretical prediction of hydrogen bonding potential surfaces, using a minimal basis to represent the atomic orbitals (STO-3G). Using this method, with a systematically chosen “correction factor”, one can compute potential surfaces with an STO-3G basis as accurately as with a much more flexible atomic basis.

167 citations

Journal ArticleDOI
TL;DR: S-layer-protein-expressing cells of strain JCM 5810 adhered to collagen-containing regions in the chicken colon, suggesting that CbsA-mediated collagen binding represents a true tissue adherence property of L. crispatus.
Abstract: The cbsA gene of Lactobacillus crispatus strain JCM 5810, encoding a protein that mediates adhesiveness to collagens, was characterized and expressed in Escherichia coli. The cbsA open reading frame encoded a signal sequence of 30 amino acids and a mature polypeptide of 410 amino acids with typical features of a bacterial S-layer protein. The cbsA gene product was expressed as a His tag fusion protein, purified by affinity chromatography, and shown to bind solubilized as well as immobilized type I and IV collagens. Three other Lactobacillus S-layer proteins, SlpA, CbsB, and SlpnB, bound collagens only weakly, and sequence comparisons of CbsA with these S-layer proteins were used to select sites in cbsA where deletions and mutations were introduced. In addition, hybrid S-layer proteins that contained the N or the C terminus from CbsA, SlpA, or SlpnB as well as N- and C-terminally truncated peptides from CbsA were constructed by gene fusion. Analysis of these molecules revealed the major collagen-binding region within the N-terminal 287 residues and a weaker type I collagen-binding region in the C terminus of the CbsA molecule. The mutated or hybrid CbsA molecules and peptides that failed to polymerize into a periodic S-layer did not bind collagens, suggesting that the crystal structure with a regular array is optimal for expression of collagen binding by CbsA. Strain JCM 5810 was found to contain another S-layer gene termed cbsB that was 44% identical in sequence to cbsA. RNA analysis showed that cbsA, but not cbsB, was transcribed under laboratory conditions. S-layer-protein-expressing cells of strain JCM 5810 adhered to collagen-containing regions in the chicken colon, suggesting that CbsA-mediated collagen binding represents a true tissue adherence property of L. crispatus.

167 citations

Journal ArticleDOI
TL;DR: In Helsinki, a Small but General Set of Manipulative Operations for Boundary Models of Solid Objects has been used to Construct a Comprehensive Solid Modeling System.
Abstract: In Helsinki, a Small but General Set of Manipulative Operations for Boundary Models of Solid Objects Has Been Used to Construct a Comprehensive Solid Modeling System.

167 citations

Proceedings Article
01 Jan 2019
TL;DR: This work enables practical deep learning while preserving benefits of Bayesian principles, and applies techniques such as batch normalisation, data augmentation, and distributed training to achieve similar performance in about the same number of epochs as the Adam optimiser.
Abstract: Bayesian methods promise to fix many shortcomings of deep learning, but they are impractical and rarely match the performance of standard methods, let alone improve them. In this paper, we demonstrate practical training of deep networks with natural-gradient variational inference. By applying techniques such as batch normalisation, data augmentation, and distributed training, we achieve similar performance in about the same number of epochs as the Adam optimiser, even on large datasets such as ImageNet. Importantly, the benefits of Bayesian principles are preserved: predictive probabilities are well-calibrated, uncertainties on out-of-distribution data are improved, and continual-learning performance is boosted. This work enables practical deep learning while preserving benefits of Bayesian principles. A PyTorch implementation is available as a plug-and-play optimiser.

167 citations

Journal ArticleDOI
TL;DR: In this article, the twisted Gaussian Schell-model (GSM) beams are interpreted in physical-optics terms by decomposition of such beams into weighted superpositions of overlapping, mutually uncorrelated but spatially coherent component fields.
Abstract: The twisted Gaussian Schell-model (GSM) beams, recently introduced by Simon and Mukunda [ J. Opt. Soc. Am. A9, 95 ( 1993)], are interpreted in physical-optics terms by decomposition of such beams into weighted superpositions of overlapping, mutually uncorrelated but spatially coherent component fields. The decomposition provides considerable physical insight into the propagation characteristics of the twisted GSM beams and also suggests convenient practical methods for generating these novel wave fields. Key properties of the twisted GSM beams are demonstrated experimentally by use of an acousto-optic coherence control technique to supply the necessary partially coherent fields.

167 citations


Authors

Showing all 8962 results

NameH-indexPapersCitations
Ashok Kumar1515654164086
Hannu Kurki-Suonio13843399607
Nicolas Gisin12582764298
Anne Lähteenmäki11648581977
Riitta Hari11149143873
Andreas Richter11076948262
Mika Sillanpää96101944260
Markku Leskelä9487636881
Ullrich Scherf9273536972
Mikko Ritala9158429934
Axel H. E. Müller8956430283
Karl Henrik Johansson88108933751
T. Poutanen8612033158
Elina Lindfors8642023846
Günter Breithardt8555433165
Network Information
Related Institutions (5)
École Polytechnique Fédérale de Lausanne
98.2K papers, 4.3M citations

95% related

Delft University of Technology
94.4K papers, 2.7M citations

95% related

Georgia Institute of Technology
119K papers, 4.6M citations

93% related

École Normale Supérieure
99.4K papers, 3M citations

93% related

Technical University of Denmark
66.3K papers, 2.4M citations

93% related

Performance
Metrics
No. of papers from the Institution in previous years
YearPapers
2021154
2020153
2019155
201851
201714
201630