Search

In radio astronomy, radio frequency interference (RFI) becomes more and more serious for radio observational facilities. The RFI always influences the search and study of the interesting astronomical objects. Mitigating the RFI becomes an essential procedure in any survey data processing. Five-hundred-meter Aperture Spherical radio Telescope (FAST) is an extremely sensitive radio telescope. It is necessary to find out an effective and precise RFI mitigation method for FAST data processing. In this work, we introduce a method to mitigate the RFI in FAST spectral observation and make a statistics for the RFI using around 300 hours FAST data. The details are as follows. Firstly, according to the characteristics of FAST spectra, we propose to use the ArPLS algorithm for baseline fitting. Our test results show that it has a good performance. Secondly, we flag the RFI with four strategies, which are to flag extremely strong RFI, flag long-lasting RFI, flag polarized RFI, and flag beam-combined RFI, respectively. The test results show that all the RFI above a preset threshold could be flagged. Thirdly, we make a statistics for the probabilities of polarized XX and YY RFI in FAST observations. The statistical results could tell us which frequencies are relatively quiescent. With such statistical data, we are able to avoid using such frequencies in our spectral observations. Finally, based on the around 300 hours FAST data, we got an RFI table, which is the most complete database currently for FAST., Comment: Accepted by Research in Astronomy and Astrophysics (RAA)

We present multi-wavelength data to investigate the physical properties and fragmentation of AFGL 333- Ridge. A statistical analysis of velocity dispersion indicates that turbulence is the dominant motion in the ridge. However, the linear mass density (1124.0 M/pc) of AFGL 333-Ridge far exceeds its critical value of 406.5 M/pc, suggesting that additional motions are required to prevent the filament radial collapse. Using the getsources algorithm, we identified 14 cores from the Herschel maps, including two protostellar cores and 12 starless cores. All of these starless cores are gravitationally bound, and are therefore considered to be prestellar cores. Based on their radius-mass relation, 11 of 14 cores have the potential to form massive stars. Moreover, the seven cores in two sub-filaments of AFGL 333-Ridge seem to constitute two necklace-like chains with a spacing length of 0.51 pc and 0.45 pc, respectively. Compared the spacing length with theoretical prediction lengths by Jeans and cylindrical fragmentations, we argued that the combination of turbulence and thermal pressure may lead to the fragmentation of the two sub-filaments into the cores., Comment: 7 pages, 4 figures, 2 tables, accepted by AJ

We present a new high-sensitivity HI observation toward nearby spiral galaxy M101 and its adjacent 2$^{\circ}\times$ 2$^{\circ}$ region using the Five-hundred-meter Aperture Spherical radio Telescope (FAST). From the observation, we detect a more extended and asymmetric HI disk around M101. While the HI velocity field within the M101's optical disk region is regular, indicating that the relatively strong disturbance occurs in its outer disk. Moreover, we identify three new HI clouds located on the southern edge of the M101's HI disk. The masses of the three HI clouds are 1.3$\times$10$^{7}$ $M_{\odot}$, 2.4$\times$10$^{7}$ $M_{\odot}$, and 2.0$\times$10$^{7}$ $M_{\odot}$, respectively. The HI clouds similar to dwarf companion NGC 5477 rotate with the HI disk of M101. Unlike the NGC 5477, they have no optical counterparts. Furthermore, we detect a new HI tail in the extended HI disk of M101. The HI tail detected gives a reliable evidence for M101 interaction with the dwarf companion NGC 5474. We argue that the extra-planar gas (three HI clouds) and the HI tail detected in the M101's disk may origin from a minor interaction with NGC 5474., Comment: 7 pages, 6 figures, accepted for publication in ApJ

In the framework of algebraic inversion, Magnetic Resonance Elastography (MRE) repeatability, reproducibility and robustness were evaluated on extracted shear velocities (or elastic moduli). The same excitation system was implemented at two sites equipped with clinical MR scanners of 1.5 T and 3 T. A set of four elastic, isotropic, homogeneous calibrated phantoms of distinct elasticity representing the spectrum of liver fibrosis severity was mechanically characterized. The repeatability of the measurements and the reproducibility between the two platforms were found to be excellent with mean coefficients of variations of 1.62% for the shear velocity mean values and 1.95% for the associated standard deviations. MRE velocities were robust to the amplitude and pattern variations of the displacement field with virtually no difference between outcomes from both magnets at identical excitation frequencies even when the displacement field amplitude was 6 times smaller. However, MRE outcomes were very sensitive to the number of voxels per wavelength, s, of the recorded displacement field, with relative biases reaching 62% and precision losing up to a factor 23.5. For both magnetic field strengths, MRE accuracy and precision were largely degraded outside of established conditions of validity ($6 \lesssim s \lesssim 9$) resulting in estimated shear velocity values not significantly different between phantoms of increasing elasticity. When fulfilling the spatial sampling conditions, either prospectively in the acquisition or retrospectively before the reconstruction, MRE produced quantitative measurements that allowed to unambiguously discriminate, with infinitesimal p-values, between the phantoms mimicking increasing severity of liver fibrosis.

Deep neural networks often suffer from poor performance or even training failure due to the ill-conditioned problem, the vanishing/exploding gradient problem, and the saddle point problem. In this paper, a novel method by acting the gradient activation function (GAF) on the gradient is proposed to handle these challenges. Intuitively, the GAF enlarges the tiny gradients and restricts the large gradient. Theoretically, this paper gives conditions that the GAF needs to meet, and on this basis, proves that the GAF alleviates the problems mentioned above. In addition, this paper proves that the convergence rate of SGD with the GAF is faster than that without the GAF under some assumptions. Furthermore, experiments on CIFAR, ImageNet, and PASCAL visual object classes confirm the GAF's effectiveness. The experimental results also demonstrate that the proposed method is able to be adopted in various deep neural networks to improve their performance. The source code is publicly available at https://github.com/LongJin-lab/Activated-Gradients-for-Deep-Neural-Networks.