High Quality Candidate Generation and Sequential Graph Attention Network for Entity Linking

Author(s):  
Zheng Fang ◽  
Yanan Cao ◽  
Ren Li ◽  
Zhenyu Zhang ◽  
Yanbing Liu ◽  
...  
Micromachines ◽  
2021 ◽  
Vol 12 (11) ◽  
pp. 1418
Author(s):  
Yue Yu ◽  
Kun She ◽  
Jinhua Liu

Medical imaging is widely used in medical diagnosis. The low-resolution image caused by high hardware cost and poor imaging technology leads to the loss of relevant features and even fine texture. Obtaining high-quality medical images plays an important role in disease diagnosis. A surge of deep learning approaches has recently demonstrated high-quality reconstruction for medical image super-resolution. In this work, we propose a light-weight wavelet frequency separation attention network for medical image super-resolution (WFSAN). WFSAN is designed with separated-path for wavelet sub-bands to predict the wavelet coefficients, considering that image data characteristics are different in the wavelet domain and spatial domain. In addition, different activation functions are selected to fit the coefficients. Inputs comprise approximate sub-bands and detail sub-bands of low-resolution wavelet coefficients. In the separated-path network, detail sub-bands, which have more sparsity, are trained to enhance high frequency information. An attention extension ghost block is designed to generate the features more efficiently. All results obtained from fusing layers are contracted to reconstruct the approximate and detail wavelet coefficients of the high-resolution image. In the end, the super-resolution results are generated by inverse wavelet transform. Experimental results show that WFSAN has competitive performance against state-of-the-art lightweight medical imaging methods in terms of quality and quantitative metrics.


Information ◽  
2019 ◽  
Vol 10 (2) ◽  
pp. 46 ◽  
Author(s):  
Chen Liu ◽  
Feng Li ◽  
Xian Sun ◽  
Hongzhe Han

Entity linking (also called entity disambiguation) aims to map the mentions in a given document to their corresponding entities in a target knowledge base. In order to build a high-quality entity linking system, efforts are made in three parts: Encoding of the entity, encoding of the mention context, and modeling the coherence among mentions. For the encoding of entity, we use long short term memory (LSTM) and a convolutional neural network (CNN) to encode the entity context and entity description, respectively. Then, we design a function to combine all the different entity information aspects, in order to generate unified, dense entity embeddings. For the encoding of mention context, unlike standard attention mechanisms which can only capture important individual words, we introduce a novel, attention mechanism-based LSTM model, which can effectively capture the important text spans around a given mention with a conditional random field (CRF) layer. In addition, we take the coherence among mentions into consideration with a Forward-Backward Algorithm, which is less time-consuming than previous methods. Our experimental results show that our model obtains a competitive, or even better, performance than state-of-the-art models across different datasets.


1966 ◽  
Vol 24 ◽  
pp. 51-52
Author(s):  
E. K. Kharadze ◽  
R. A. Bartaya

The unique 70-cm meniscus-type telescope of the Abastumani Astrophysical Observatory supplied with two objective prisms and the seeing conditions characteristic at Mount Kanobili (Abastumani) permit us to obtain stellar spectra of a high quality. No additional design to improve the “climate” immediately around the telescope itself is being applied. The dispersions and photographic magnitude limits are 160 and 660Å/mm, and 12–13, respectively. The short-wave end of spectra reaches 3500–3400Å.


Author(s):  
R. L. Lyles ◽  
S. J. Rothman ◽  
W. Jäger

Standard techniques of electropolishing silver and silver alloys for electron microscopy in most instances have relied on various CN recipes. These methods have been characteristically unsatisfactory due to difficulties in obtaining large electron transparent areas, reproducible results, adequate solution lifetimes, and contamination free sample surfaces. In addition, there are the inherent health hazards associated with the use of CN solutions. Various attempts to develop noncyanic methods of electropolishing specimens for electron microscopy have not been successful in that the specimen quality problems encountered with the CN solutions have also existed in the previously proposed non-cyanic methods.The technique we describe allows us to jet polish high quality silver and silver alloy microscope specimens with consistant reproducibility and without the use of CN salts.The solution is similar to that suggested by Myschoyaev et al. It consists, in order of mixing, 115ml glacial actic acid (CH3CO2H, specific wt 1.04 g/ml), 43ml sulphuric acid (H2SO4, specific wt. g/ml), 350 ml anhydrous methyl alcohol, and 77 g thiourea (NH2CSNH2).


Author(s):  
A. V. Crewe ◽  
J. Wall ◽  
L. M. Welter

A scanning microscope using a field emission source has been described elsewhere. This microscope has now been improved by replacing the single magnetic lens with a high quality lens of the type described by Ruska. This lens has a focal length of 1 mm and a spherical aberration coefficient of 0.5 mm. The final spot size, and therefore the microscope resolution, is limited by the aberration of this lens to about 6 Å.The lens has been constructed very carefully, maintaining a tolerance of + 1 μ on all critical surfaces. The gun is prealigned on the lens to form a compact unit. The only mechanical adjustments are those which control the specimen and the tip positions. The microscope can be used in two modes. With the lens off and the gun focused on the specimen, the resolution is 250 Å over an undistorted field of view of 2 mm. With the lens on,the resolution is 20 Å or better over a field of view of 40 microns. The magnification can be accurately varied by attenuating the raster current.


Author(s):  
L. Mulestagno ◽  
J.C. Holzer ◽  
P. Fraundorf

Due to the wealth of information, both analytical and structural that can be obtained from it TEM always has been a favorite tool for the analysis of process-induced defects in semiconductor wafers. The only major disadvantage has always been, that the volume under study in the TEM is relatively small, making it difficult to locate low density defects, and sample preparation is a somewhat lengthy procedure. This problem has been somewhat alleviated by the availability of efficient low angle milling.Using a PIPS® variable angle ion -mill, manufactured by Gatan, we have been consistently obtaining planar specimens with a high quality thin area in excess of 5 × 104 μm2 in about half an hour (milling time), which has made it possible to locate defects at lower densities, or, for defects of relatively high density, obtain information which is statistically more significant (table 1).


Author(s):  
C. O. Jung ◽  
S. J. Krause ◽  
S.R. Wilson

Silicon-on-insulator (SOI) structures have excellent potential for future use in radiation hardened and high speed integrated circuits. For device fabrication in SOI material a high quality superficial Si layer above a buried oxide layer is required. Recently, Celler et al. reported that post-implantation annealing of oxygen implanted SOI at very high temperatures would eliminate virtually all defects and precipiates in the superficial Si layer. In this work we are reporting on the effect of three different post implantation annealing cycles on the structure of oxygen implanted SOI samples which were implanted under the same conditions.


Author(s):  
Judith M. Brock ◽  
Max T. Otten ◽  
Marc. J.C. de Jong

A Field Emission Gun (FEG) on a TEM/STEM instrument provides a major improvement in performance relative to one equipped with a LaB6 emitter. The improvement is particularly notable for small-probe techniques: EDX and EELS microanalysis, convergent beam diffraction and scanning. The high brightness of the FEG (108 to 109 A/cm2srad), compared with that of LaB6 (∼106), makes it possible to achieve high probe currents (∼1 nA) in probes of about 1 nm, whilst the currents for similar probes with LaB6 are about 100 to 500x lower. Accordingly the small, high-intensity FEG probes make it possible, e.g., to analyse precipitates and monolayer amounts of segregation on grain boundaries in metals or ceramics (Fig. 1); obtain high-quality convergent beam patterns from heavily dislocated materials; reliably detect 1 nm immuno-gold labels in biological specimens; and perform EDX mapping at nm-scale resolution even in difficult specimens like biological tissue.The high brightness and small energy spread of the FEG also bring an advantage in high-resolution imaging by significantly improving both spatial and temporal coherence.


Sign in / Sign up

Export Citation Format

Share Document