Search

A coupled dark energy model is considered, in which dark energy is represented by a generalized three-form field and dark matter by dust. By assuming the functions $N$ and $I$ in the model's Lagrangian as two power-law functions of the three-form field, we obtain two fixed points of the autonomous system of evolution equations, consisting of a attractor and a tracking saddle point which can be used to alleviate the coincidence problem. After marginalizing the present three-form field $\kappa X_{0}$ which is unable to be strictly restricted, we confront the model with the latest Type Ia Supernova (SN \uppercase\expandafter{\romannumeral1}a), Baryon Acoustic Oscillations (BAO) and Cosmic Microwave Backround (CMB) radiation observations with the fitting results $\Omega_{m0}= 0.280_{-0.048}^{+0.048}$ and $\lambda=0.011_{-0.032}^{+0.032}$ in the $2\sigma$ confidence level, we also find that the best fitting effective dark energy equation of state (EOS) crosses $ -1$ at redshift around 0.2., Comment: 13 pages, 13 figures, 2 tables, accepted by Physics of the Dark Universe(Volume 27, January 2020, 100462). arXiv admin note: substantial text overlap with arXiv:1704.05772

We propose a new coupled three-form dark energy model to relieve the Hubble tension in this paper. Firstly, by performing a dynamical analysis with the coupled three-form dark energy model, we obtain four fixed points, including a saddle point representing a radiation dominated Universe, a saddle point representing a matter dominated Universe, and two attractors representing two saturated de Sitter Universes. Secondly, by confronting the coupled three-form dark energy model and the $\Lambda$ cold dark matter model (the $\Lambda$CDM model) with cosmic microwave background (CMB), baryonic acoustic oscillations (BAO), Type Ia supernovae (SN Ia) observations, we obtain $H_0= 67.8_{-0.6}^{+0.7}$($1\sigma$ level) km/s/Mpc for the coupled three-form dark energy model and $H_0=67.6_{-0.5}^{+0.5}$($1\sigma$ level) km/s/Mpc for the $\Lambda$CDM model, the former is in strong tension with the latest local measured $H_0$ value at $4.3\sigma$ confidence level, while the latter is in strong tension with the latest local measured $H_0$ value at $5.1\sigma$ level., Comment: 16 pages, 1 figure, 3 table. arXiv admin note: substantial text overlap with arXiv:2011.09160

The backreaction term ${\cal Q}_\CD$ and the averaged spatial Ricci scalar $\average{\CR}$ in the spatially averaged inhomogeneous Universe can be used to combine into effective perfect fluid energy density $\varrho_{\rm eff}^{\CD}$ and pressure $p_{\rm eff}^{\CD}$ that can be regarded as new effective sources for the backreaction effects. In order to model the realistic evolution of backreaction, we adopt the Chevallier-Polarski-Linder(CPL) parameterizations of the equation of state(EoS) of the effective perfect fluid. To deal with observations in the backreaction context, in this paper, we employ two metrics to describe the the late time Universe, one of them is the standard Friedmann-Lema\^{\i}tre-Robertson-Walker(FLRW) metric, and the other is a template metric with an evolving curvature parameter introduced by Larena et. al. in \cite{larena2009testing}. We also fit the CPL backreaction model using type Ia supernova(SN Ia) data and observational Hubble parameter data(OHD) with these two metrics, and find out that parameter tensions between two different data sets are larger when the backreaction model is equipped with the template metric, therefore we conclude that the prescription of the geometrical instantaneous spatially-constant curvature $\kappa_{\CD}$ needs to be modified., Comment: 16 pages, 6 figures, 3 table. arXiv admin note: substantial text overlap with arXiv:1707.00111

Rearrangement tasks have been identified as a crucial challenge for intelligent robotic manipulation, but few methods allow for precise construction of unseen structures. We propose a visual foresight model for pick-and-place rearrangement manipulation which is able to learn efficiently. In addition, we develop a multi-modal action proposal module which builds on the Goal-Conditioned Transporter Network, a state-of-the-art imitation learning method. Our image-based task planning method, Transporters with Visual Foresight, is able to learn from only a handful of data and generalize to multiple unseen tasks in a zero-shot manner. TVF is able to improve the performance of a state-of-the-art imitation learning method on unseen tasks in simulation and real robot experiments. In particular, the average success rate on unseen tasks improves from 55.4% to 78.5% in simulation experiments and from 30% to 63.3% in real robot experiments when given only tens of expert demonstrations. Video and code are available on our project website: https://chirikjianlab.github.io/tvf/, Comment: IEEE IROS 2022

Recently, Vision Transformer (ViT) has continuously established new milestones in the computer vision field, while the high computation and memory cost makes its propagation in industrial production difficult. Pruning, a traditional model compression paradigm for hardware efficiency, has been widely applied in various DNN structures. Nevertheless, it stays ambiguous on how to perform exclusive pruning on the ViT structure. Considering three key points: the structural characteristics, the internal data pattern of ViTs, and the related edge device deployment, we leverage the input token sparsity and propose a computation-aware soft pruning framework, which can be set up on vanilla Transformers of both flatten and CNN-type structures, such as Pooling-based ViT (PiT). More concretely, we design a dynamic attention-based multi-head token selector, which is a lightweight module for adaptive instance-wise token selection. We further introduce a soft pruning technique, which integrates the less informative tokens generated by the selector module into a package token that will participate in subsequent calculations rather than being completely discarded. Our framework is bound to the trade-off between accuracy and computation constraints of specific edge devices through our proposed computation-aware training strategy. Experimental results show that our framework significantly reduces the computation cost of ViTs while maintaining comparable performance on image classification. Moreover, our framework can guarantee the identified model to meet resource specifications of mobile devices and FPGA, and even achieve the real-time execution of DeiT-T on mobile platforms. For example, our method reduces the latency of DeiT-T to 26 ms (26%$\sim $41% superior to existing works) on the mobile device with 0.25%$\sim $4% higher top-1 accuracy on ImageNet., Comment: ECCV 2022

We firstly derive the stellar population properties: age and metallicity for $\sim$ 43,000 low redshift galaxies in the seventh data release (DR7) of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) survey, which have no spectroscopic observations in the Sloan Digital Sky Survey(SDSS). We employ a fitting procedure based on the small-scale features of galaxy spectra so as to avoid possible biases from the uncertain flux calibration of the LAMOST spectroscopy. We show that our algorithm can successfully recover the average age and metallicity of the stellar populations of galaxies down to signal-to-noise$\geq$5 through testing on both mock galaxies and real galaxies comprising LAMOST and their SDSS counterparts. We provide a catalogue of the age and metallicity for $\sim$ 43,000 LAMOST galaxies online. As a demonstration of the scientific application of this catalogue, we present the Holmberg effect on both age and metallicity of a sample of galaxies in galaxy pairs., Comment: 15 pages, 9 Postscript figures, accepted for publication in The Astrophysical Journal Supplement Series