scispace - formally typeset
Search or ask a question
Author

Xiang Zhang

Bio: Xiang Zhang is an academic researcher from Baylor College of Medicine. The author has contributed to research in topics: Medicine & Computer science. The author has an hindex of 154, co-authored 1733 publications receiving 117576 citations. Previous affiliations of Xiang Zhang include University of California, Berkeley & University of Texas MD Anderson Cancer Center.


Papers
More filters
Journal ArticleDOI
TL;DR: The ADSL and LPL gene mutations are correlated with differences in meat quality in different chicken breeds, and high-resolution melting curve is an effective prediction technology for these mutations.
Abstract: Adenylosuccinate lyase (ADSL) and lipoprotein lipase (LPL) are key enzymes in the metabolism of inosine monophosphate (IMP) and fat mass, which are important factors in meat quality evaluation. In this study, we selected 50 hens from the ISA B-line layers and Guangxi Yellow chickens, slaughtered the chickens at 120 days old, and analyzed polymorphisms in the ADSL and LPL genes using the high-resolution melting curve method. Blood lipid parameters, intramuscular fat (IMF), and IMP content were higher (P < 0.05) in Guangxi Yellow chickens than in ISA B-line layers, while LPL activity was lower (P < 0.05). In exon 2 of the ADSL gene, a C3484T mutation was identified. In both breeds, the CC genotype showed the highest IMP, and IMP was the lowest in the TT genotype. In the 5ꞌ regulatory region of the LPL gene, a C293T mutation was identified. In both breeds, the CC genotype showed the lowest LPL and IMF, while IMF was the highest in the TT genotype. The percentages of individuals with the TT type in the ADSL gene, which was associated with the lowest IMP, were 16.0 and 52.0% in Guangxi chickens and ISA layers, respectively. The percentages of individuals with the CC type of the LPL gene, which was associated with the lowest LPL and IMF, were 28.0 and 44.0%, respectively. The ADSL and LPL gene mutations are correlated with differences in meat quality in different chicken breeds, and high-resolution melting curve is an effective prediction technology for these mutations.

14 citations

Journal ArticleDOI
Jaroslav Adam1, Dagmar Adamová2, Madan M. Aggarwal3, G. Aglieri Rinella4  +1041 moreInstitutions (100)
TL;DR: F Fourier decompositions of P_{2}, studied as a function of the collision centrality, show that correlations at |Δη|≥0.9 can be well reproduced by a flow ansatz based on the notion that measured transverse momentum correlations are strictly determined by the collective motion of the system.
Abstract: We present the first measurement of the two-particle transverse momentum differential correlation function, P2≡⟨ΔpTΔpT⟩/⟨pT⟩2, in Pb-Pb collisions at √sNN=2.76 TeV. Results for P2 are reported as a function of the relative pseudorapidity (Δη) and azimuthal angle (Δφ) between two particles for different collision centralities. The Δϕ dependence is found to be largely independent of Δη for |Δη|≥0.9. In the 5% most central Pb-Pb collisions, the two-particle transverse momentum correlation function exhibits a clear double-hump structure around Δφ=π (i.e., on the away side), which is not observed in number correlations in the same centrality range, and thus provides an indication of the dominance of triangular flow in this collision centrality. Fourier decompositions of P2, studied as a function of the collision centrality, show that correlations at |Δη|≥0.9 can be well reproduced by a flow ansatz based on the notion that measured transverse momentum correlations are strictly determined by the collective motion of the system.

14 citations

Journal ArticleDOI
TL;DR: In this paper, a polycrystalline ceramic is prepared via sol-gel process and its magnetic properties and electron spin resonance (ESR) spectra have been investigated experimentally.

14 citations

Journal ArticleDOI
TL;DR: In this paper, a method to determine the adhesion force of polymeric microstructures fabricated by microstereolithography (μSL) is presented. But the method is limited to the case of 1,6-hexanediol diacrylate (HDDA) parallel beams.
Abstract: The adhesion between microstructures represents a great challenge in reliability of polymeric three-dimensional structures fabricated by microstereolithography (μSL). During the evaporative releasing, the capillary force of the solvent causes the deformation and adhesion of the fabricated beams. We present a method to determine the adhesion force of polymeric microstructures fabricated by μSL. The test structures with parallel beams were fabricated and released from the liquid resin via evaporation. By measuring the relationship between the adhesion length and the geometry of the beams, the adhesion force between two 1,6-hexanediol diacrylate (HDDA) polymeric parallel beams is determined as γ=72±5 mN/m. This simple method and the determined adhesion force provide a key in designing reliable polymeric microelectromechanical systems in preventing the stiction problem.

14 citations

Book ChapterDOI
01 Jan 2015
TL;DR: This paper proposes a methodology to evaluate OSM roads that does not rely on reference data or ground truth, and identifies several issues in tagging and modeling OSM road network by case studies, and also gives suggestions for contributors and routing service providers.
Abstract: OpenStreetMap (OSM) is mostly a good map for viewers to look at but it lacks of sufficient quality in certain applications like navigation. Quality issues are usually related to how roads are ‘drawn’ (modeled) by OSM contributors. First of all, this paper identifies several issues in tagging and modeling OSM road network by case studies, and also gives suggestions for contributors and routing service providers. As a key contribution, this paper proposes a methodology to evaluate OSM roads that does not rely on reference data or ground truth. The evaluation aims not only to identify errors in OSM data, but also to give more intelligent suggestions based on the information available in the spatial context of the problematic data. More specifically, named roads are recognized based on the concept of “stroke”. Missing or incorrect names can be found by outlier detection within the scope of the named roads. Such an idea can be widely applied to detect inconsistent tags and provide intelligent suggestions for data correction.

14 citations


Cited by
More filters
Proceedings ArticleDOI
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

123,388 citations

Proceedings Article
04 Sep 2014
TL;DR: This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

55,235 citations

Journal ArticleDOI
04 Mar 2011-Cell
TL;DR: Recognition of the widespread applicability of these concepts will increasingly affect the development of new means to treat human cancer.

51,099 citations

Posted Content
TL;DR: This work presents a residual learning framework to ease the training of networks that are substantially deeper than those used previously, and provides comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

44,703 citations

Proceedings ArticleDOI
07 Jun 2015
TL;DR: Inception as mentioned in this paper is a deep convolutional neural network architecture that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14).
Abstract: We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

40,257 citations