University of California

About: University of California is a education organization based out in Oakland, California, United States. It is known for research contribution in the topics: Population & Layer (electronics). The organization has 55175 authors who have published 52933 publications receiving 1491169 citations. The organization is also known as: UC & University of California System.

CREST: Convolutional Residual Learning for Visual Tracking

Abstract: Discriminative correlation filters (DCFs) have been shown to perform superiorly in visual tracking. They only need a small set of training samples from the initial frame to generate an appearance model. However, existing DCFs learn the filters separately from feature extraction, and update these filters using a moving average operation with an empirical weight. These DCF trackers hardly benefit from the end-to-end training. In this paper, we propose the CREST algorithm to reformulate DCFs as a one-layer convolutional neural network. Our method integrates feature extraction, response map generation as well as model update into the neural networks for an end-to-end training. To reduce model degradation during online update, we apply residual learning to take appearance changes into account. Extensive experiments on the benchmark datasets demonstrate that our CREST tracker performs favorably against state-of-the-art trackers.

Inverse Radiative Heat Transfer

Abstract: Up to this point we have concerned ourselves with radiative heat transfer problems, where the necessary geometry, temperatures, and radiative properties are known, enabling us to calculate the radiative intensity and radiative heat fluxes in such enclosures. Such cases are sometimes called “direct” heat transfer problems. However, there are many important engineering applications where knowledge of one or more input parameters is desired that cause a certain radiative intensity field. For example, it may be desired to control the temperatures of heating elements in a furnace, in order to achieve a specified temperature distribution or radiative heat load on an object being heated. Or the aim may be to deduce difficult to measure parameters (such as radiative properties, temperature fields inside a furnace, etc.) based on measurements of radiative intensity or radiative flux. Such calculations are known as inverse heat transfer analyses.

Pathological classification and molecular genetics of meningiomas

Abstract: Meningiomas are extremely common adult brain tumors originating from meningeal coverings of the brain and spinal cord. While most are slowly growing Word Health organization (WHO) grade I tumors, rare variants (clear cell, chordoid, papillary, and rhabdoid), as well as brain invasive (WHO grade II), atypical (WHO grade II), and anaplastic (WHO grade III) meningiomas are considerably more aggressive. This review summarizes the histopathological and genetic features of meningiomas, including differential diagnosis, pitfalls, and grading challenges. Early stages of meningioma tumorigenesis are closely linked to inactivation of one or more members of the 4.1 superfamily, including the neurofibromatosis type 2 (NF2) and 4.1B (DAL-1) genes, which interact with the 14-3-3 protein family. Other chromosome 22q genes implicated include BAM22, BCR (breakpoint cluster region), and TIMP-1, the last of which is implicated in higher-grade meningiomas. Atypical meningiomas also commonly show chromosomal losses of 1p, 6q, 10, 14q, and 18q, as well as multiple chromosomal gains. While most relevant genes remain unknown, two chromosome 14q candidates (MEG3 and NDRG2) have recently been identified. In addition to alterations of CDKN2A, p14 ARF , and CDKN2B tumor suppressor genes on 9p21, a contribution of the wingless (wnt) pathway with alterations of the E-cadherin and beta-catenin proteins, as well as alterations of the hedgehog signaling pathway have been implicated in anaplastic meningiomas. The integration of histopathological appearance, complex genetic/genomic data, and outcome will likely result in the identification of clinically distinct meningioma subgroups, which in turn can facilitate the development of targeted therapeutic strategies.

Physical activity for the chronically ill and disabled.

Abstract: Exercise prescription principles for persons without chronic disease and/or disability are based on well developed scientific information. While there are varied objectives for being physically active, including enhancing physical fitness, promoting health by reducing the risk for chronic disease and ensuring safety during exercise participation, the essence of the exercise prescription is based on individual interests, health needs and clinical status, and therefore the aforementioned goals do not always carry equal weight. In the same manner, the principles of exercise prescription for persons with chronic disease and/or disability should place more emphasis on the patient's clinical status and, as a result, the exercise mode, intensity, frequency and duration are usually modified according to their clinical condition. Presently, these exercise prescription principles have been scientifically defined for clients with coronary heart disease. However, other diseases and/or disabilities have been studied less (e.g. renal failure, cancer, chronic fatigue syndrome, cerebral palsy). This article reviews these issues with specific reference to persons with chronic diseases and disabilities.

Protein fold recognition using sequence-derived predictions

Abstract: A computer-assisted method for assigning an amino acid probe sequence to a known three-dimensional protein structure. In particular, the invention includes a method for using the amino acid sequence of a probe plus sequence-derived properties of the probe in making fold assignments. The method includes inputting into a computer system a string p 1 , p 2 . . . p n describing the amino acid sequence of the probe sequence and at least one sequence-derived property for the probe sequence; inputting into a computer system a string t 1 , t 2 . . . t m of structural properties for each member of a library of known 3D structures; executing an alignment algorithm in the computer to compute an alignment score indicating the optimal alignment of the string p 1 , p 2 . . . p n to each string t 1 , t 2 . . . t m by applying a combined compatibility function g(p i , t j ); determining the statistical significance of each alignment score to determine a best-fit alignment score; and applying the best-fit alignment score to indicate or select the corresponding 3D protein structure from the library for output to a user. The invention includes a computer program implementation of such a method.