arXiv subjects

Siyu Li

Publications and source records attributed to Siyu Li.

At least 73 records · Page 4Linked to original sources

arXiv2023

An End-to-End Human Simulator for Task-Oriented Multimodal Human-Robot Collaboration

This paper proposes a neural network-based user simulator that can provide a multimodal interactive environment for training Reinforcement Learning (RL) agents in collaborative tasks involving multiple modes of communication. The simulator is trained on the existing ELDERLY-AT-HOME corpus and accommodates multiple modalities such as language, pointing gestures, and haptic-ostensive actions. The paper also presents a novel multimodal data augmentation approach, which addresses the challenge of using a limited dataset due to the expensive and time-consuming nature of collecting human demonstrations. Overall, the study highlights the potential for using RL and multimodal user simulators in developing and improving domestic assistive robots.

cs.RO↗

arXiv2023

Forecasts of CMB lensing reconstruction of AliCPT-1 from the foreground cleaned polarization data

Cosmic microwave background radiation (CMB) observations are unavoidably contaminated by emission from various extra-galactic foregrounds, which must be removed to obtain reliable measurements of the cosmological signal. In this paper, we demonstrate CMB lensing reconstruction in AliCPT-1 after foreground removal, combine the two bands of AliCPT-1 (90 and 150~GHz) with Planck HFI bands (100, 143, 217 and 353~GHz) and with the WMAP-K band (23~GHz). In order to balance contamination by instrumental noise and foreground residual bias, we adopt the Needlet Internal Linear Combination (NILC) method to clean the E-map and the constrained Internal Linear Combination (cILC) method to clean the B-map. The latter utilizes additional constraints on average frequency scaling of the dust and synchrotron to remove foregrounds at the expense of somewhat noisier maps. Assuming 4 modules observing 1 season from simulation data, the resulting effective residual noise in E- and B-map are roughly $15~μ{\rm K}\cdot{\rm arcmin}$ and $25~μ{\rm K}\cdot{\rm arcmin}$, respectively. As a result, the CMB lensing reconstruction signal-to-noise ratio (SNR) from polarization data is about SNR$\,\approx\,$4.5. This lensing reconstruction capability is comparable to that of other stage-III small aperture millimeter CMB telescopes.

astro-ph.CO↗

arXiv2023

A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPT

Recently, ChatGPT, along with DALL-E-2 and Codex,has been gaining significant attention from society. As a result, many individuals have become interested in related resources and are seeking to uncover the background and secrets behind its impressive performance. In fact, ChatGPT and other Generative AI (GAI) techniques belong to the category of Artificial Intelligence Generated Content (AIGC), which involves the creation of digital content, such as images, music, and natural language, through AI models. The goal of AIGC is to make the content creation process more efficient and accessible, allowing for the production of high-quality content at a faster pace. AIGC is achieved by extracting and understanding intent information from instructions provided by human, and generating the content according to its knowledge and the intent information. In recent years, large-scale models have become increasingly important in AIGC as they provide better intent extraction and thus, improved generation results. With the growth of data and the size of the models, the distribution that the model can learn becomes more comprehensive and closer to reality, leading to more realistic and high-quality content generation. This survey provides a comprehensive review on the history of generative models, and basic components, recent advances in AIGC from unimodal interaction and multimodal interaction. From the perspective of unimodality, we introduce the generation tasks and relative models of text and image. From the perspective of multimodality, we introduce the cross-application between the modalities mentioned above. Finally, we discuss the existing open problems and future challenges in AIGC.

cs.AI↗

arXiv2022

Accelerating spherical harmonic transforms for a large number of sky maps

The spherical harmonic transform is a powerful tool in the analysis of spherical data sets, such as the cosmic microwave background data. In this work, we present a new scheme for the spherical harmonic transforms that supports both CPU and GPU computations, which is specially efficient on a large number of sky maps. By comparing our implementation with the standard Libsharp-HEALPix program, we demonstrate 2-10 times speedup for the CPU implementation, and up to 30 times speedup when a state-of-the-art GPU is employed. This new scheme's software package is available via an open source GitHub repository.

astro-ph.IM↗

arXiv2022

Performance forecasts for the primordial gravitational wave detection pipelines for AliCPT-1

AliCPT is the first Chinese cosmic microwave background (CMB) experiment which will make the most precise measurements of the CMB polarization in the northern hemisphere. The key science goal for AliCPT is the detection of primordial gravitational waves (PGWs). It is well known that an epoch of cosmic inflation, in the very early universe, can produce PGWs, which leave an imprint on the CMB in form of odd parity $B$-mode polarization. In this work, we study the performance of the component separation and parameter estimation pipelines in context of constraining the value of the tensor-to-scalar ratio. Based on the simulated data for one observation season, we compare five different pipelines with different working principles. Three pipelines perform component separation at map or spectra level before estimating $r$ from the cleaned spectra, while the other two pipelines performs a global fit for both foreground parameters and $r$. We also test different methods to account for the effects of time stream filtering systematics. This work shows that our pipelines provide consistent and robust constraints on the tensor-to-scalar ratio and a consistent sensitivity $σ(r) \sim 0.02$. This showcases the potential of precise $B$-mode polarization measurement with AliCPT-1. AliCPT will provide a powerful opportunity to detect PGWs, which is complementary with various ground-based CMB experiments in the southern hemisphere.

astro-ph.CO↗

arXiv2022

Biophysical Modeling of SARS-CoV-2 Assembly: Genome Condensation and Budding

The COVID-19 pandemic caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has spurred unprecedented and concerted worldwide research to curtail and eradicate this pathogen. SARS-CoV-2 has four structural proteins: Envelope (E), Membrane (M), Nucleocapsid (N), and Spike (S), which self-assemble along with its RNA into the infectious virus by budding from intracellular lipid membranes. In this paper, we develop a model to explore the mechanisms of RNA condensation by structural proteins, protein oligomerization and cellular membrane-protein interactions that control the budding process and the ultimate virus structure. Using molecular dynamics simulations, we have deciphered how the positively charged N proteins interact and condense the very long genomic RNA resulting in its packaging by a lipid envelope decorated with structural proteins inside a host cell. Furthermore, considering the length of RNA and the size of the virus, we find that the intrinsic curvature of M proteins is essential for virus budding. While most current research has focused on the S protein, which is responsible for viral entry, and it has been motivated by the need to develop efficacious vaccines, the development of resistance through mutations in this crucial protein makes it essential to elucidate the details of the viral life cycle to identify other drug targets for future therapy. Our simulations will provide insight into the viral life cycle through the assembly of viral particles de novo and potentially identify therapeutic targets for future drug development.

physics.bio-ph↗

arXiv2022

Group-based control of large-scale micro-robot swarms with on-board Physical Finite-State Machines

An important problem in microrobotics is how to control a large group of microrobots with a global control signal. This paper focuses on controlling a large-scale swarm of MicroStressBots with on-board physical finite-state machines. We introduce the concept of group-based control, which makes it possible to scale up the swarm size while reducing the complexity both of robot fabrication as well as swarm control. We prove that the group-based control system is locally accessible in terms of the robot positions. We further hypothesize based on extensive simulations that the system is globally controllable. A nonlinear optimization strategy is proposed to control the swarm by minimizing control effort. We also propose a probabilistically complete collision avoidance method that is suitable for online use. The paper concludes with an evaluation of the proposed methods in simulations.

cs.RO↗

arXiv2022

Forecasts on CMB lensing observations with AliCPT-1

AliCPT-1 is the first Chinese CMB experiment aiming for high precision measurement of Cosmic Microwave Background B-mode polarization. The telescope, currently under deployment in Tibet, will observe in two frequency bands centered at 90 and 150 GHz. We forecast the CMB lensing reconstruction, lensing-galaxy as well as lensing-CIB (Cosmic Infrared Background) cross correlation signal-to-noise ratio (SNR) for AliCPT-1. We consider two stages with different integrated observation time, namely "4 module*yr" (first stage) and "48 module*yr" (final stage). For lensing reconstruction, we use three different quadratic estimators, namely temperature-only, polarization-only and minimum-variance estimators, using curved sky geometry. We take into account the impact of inhomogeneous hit counts as well as of the mean-field bias due to incomplete sky coverage. In the first stage, our results show that the 150 GHz channel is able to measure the lensing signal at $15σ$ significance with the minimum-variance estimator. In the final stage, the measurement significance will increase to $31σ$. We also combine the two frequency data in the harmonic domain to optimize the SNR. Our result show that the coadding procedure can significantly reduce the reconstruction bias in the multiple range l>800. Thanks to the high quality of the polarization data in the final stage of AliCPT-1, the EB estimator will dominate the lensing reconstruction in this stage. We also estimate the SNR of cross-correlations between AliCPT-1 CMB lensing and other tracers of the large scale structure of the universe. For its cross-correlation with DESI galaxies/quasars, we report the cross-correlation SNR = 10-20 for the 4 redshift bins at 0.05<z<2.1. In the first stage, the total SNR is about $32$. In the final stage, the lensing-galaxy cross-correlation can reach SNR=52.

astro-ph.CO↗

arXiv2022

Virus Assembly Pathways inside a Host Cell

Simple RNA viruses self-assemble spontaneously and encapsulate their genome into a shell called the capsid. This process is mainly driven by the attractive electrostatic interaction between the positive charges on capsid proteins and the negative charges on the genome. Despite its importance and many decades of intense research, how the virus selects and packages its native RNA inside the crowded environment of a host cell cytoplasm in the presence of an abundance of non-viral RNA and other anionic polymers, has remained a mystery. In this paper, we perform a series of simulations to monitor the growth of viral shells and find the mechanism by which cargo-coat protein interactions can impact the structure and stability of the viral shells. We show that coat protein subunits can assemble around a globular nucleic acid core by forming non-icosahedral cages, which have been recently observed in assembly experiments involving small pieces of RNA. We find that the resulting cages are strained and can easily be split into fragments along stress lines. This suggests that such metastable non-icosahedral intermediates could be easily re-assembled into the stable native icosahedral shells if the larger wild-type genome becomes available, despite the presence of myriad of non-viral RNAs.

physics.bio-ph↗

arXiv2021

The impact of data volume on performance of deep learning based building rooftop extraction using very high spatial resolution aerial images

Building rooftop data are of importance in several urban applications and in natural disaster management. In contrast to traditional surveying and mapping, by using high spatial resolution aerial images, deep learning-based building rooftops extraction methods are efficient and accurate. Although more training data is preferred in deep learning-based tasks, the effect of data volume on building extraction models is underexplored. Therefore, the paper explores the impact of data volume on the performance of building rooftop extraction from very-high-spatial-resolution (VHSR) images using deep learning-based methods. To do so, we manually labelled 0.12m spatial resolution aerial images and perform a comparative analysis of models trained on datasets of different sizes using popular deep learning architectures for segmentation tasks, including Fully Convolutional Networks (FCN)-8s, U-Net and DeepLabv3+. The experiments showed that with more training data, algorithms converged faster and achieved higher accuracy, while better algorithms were able to better mitigate the lack of training data.

cs.CV↗

arXiv2021

Fast scalar quadratic maximum likelihood estimators for the CMB B-mode power spectrum

Constructing a fast and efficient estimator for the B-mode power spectrum of cosmic microwave background (CMB) is of critical importance for CMB science. For a general CMB survey, the Quadratic Maximum Likelihood (QML) estimator for CMB polarization has been proved to be the optimal estimator with minimal uncertainties, but it is computationally very expensive. In this article, we propose two new QML methods for B-mode power spectrum estimation. We use the Smith-Zaldarriaga approach to prepare pure-B mode map, and E-mode recycling method to obtain a leakage free B-mode map. We then use the scalar QML estimator to analyze the scalar pure-B map (QML-SZ) or B-mode map (QML-TC). The QML-SZ and QML-TC estimators have similar error bars as the standard QML estimators but their computational cost is nearly one order of magnitude smaller. The basic idea is that one can construct the pure B-mode CMB map by using the E-B separation method proposed by Smith-Zaldarriaga (SZ) or the one considering the template cleaning (TC) technique, then apply QML estimator to these scalar fields. By simulating potential observations of space-based and ground-based detectors, we test the reliability of these estimators by comparing them with the corresponding results of the traditional QML estimator and the pure B-mode pseudo-Cl estimator.

astro-ph.CO↗

arXiv2020

Effect of the charge distribution of virus coat proteins on the length of packaged RNAs

Single-stranded RNA viruses efficiently encapsulate their genome into a protein shell called the capsid. Electrostatic interactions between the positive charges in the capsid protein's N-terminal tail and the negatively charged genome have been postulated as the main driving force for virus assembly. Recent experimental results indicate that the N-terminal tail with the same number of charges and same lengths package different amounts of RNA, which reveals that electrostatics alone cannot explain all the observed outcomes of the RNA self-assembly experiments. Using a mean-field theory, we show that the combined effect of genome configurational entropy and electrostatics can explain to some extent the amount of packaged RNA with mutant proteins where the location and number of charges on the tails are altered. Understanding the factors contributing to the virus assembly could promote the attempt to block viral infections or to build capsids for gene therapy applications.

physics.bio-ph↗

arXiv2020

How a Virus Circumvents Energy Barriers to Form Symmetric Shells

Previous self-assembly experiments on a model icosahedral plant virus have shown that, under physiological conditions, capsid proteins initially bind to the genome through an en masse mechanism and form nucleoprotein complexes in a disordered state, which raises the questions as to how virions are assembled into a highly ordered structure in the host cell. Using small-angle X-ray scattering, we find out that a disorder-order transition occurs under physiological conditions upon an increase in capsid protein concentrations. Our cryo-transmission electron microscopy reveals closed spherical shells containing in vitro transcribed viral RNA even at pH 7.5, in marked contrast with the previous observations. We use Monte Carlo simulations to explain this disorder-order transition and find that, as the shell grows, the structures of disordered intermediates in which the distribution of pentamers does not belong to the icosahedral subgroups become energetically so unfavorable that the caps can easily dissociate and reassemble overcoming the energy barriers for the formation of perfect icosahedral shells. In addition, we monitor the growth of capsids under the condition that the nucleation and growth is the dominant pathway and show that the key for the disorder-order transition in both en masse and nucleation and growth pathways lies in the strength of elastic energy compared to the other forces in the system including protein-protein interactions and the chemical potential of free subunits. Our findings explain, at least in part, why perfect virions with icosahedral order form under different conditions including physiological ones.

physics.bio-ph↗

arXiv2019

Ground States of Crystalline Caps: Generalized Jellium on Curved Space

We study the structure and elastic energy of the ground states of crystalline caps conforming to a spherical surface. These ground states consist of positive disclination defects in structures spanning from flat and weakly curved crystals to closed shells. We compare two different continuum theories and one discrete-lattice approach for the elastic energy of defective caps. We first investigate the transition between defect-free caps to single-disclination ground states and show that it is characterized by continuous symmetry-breaking transition. Further, we show that ground states with icosahedral subgroup symmetries in caps arise across a range of curvatures, even far from the closure point of complete shells. Thus, while superficially similar to other models of 2D `jellium' (e.g. superconducting disks and 2D Wigner crystals), the interplay between the free edge of crystalline caps and the non-Euclidean geometry of its embedding leads to non trivial ground state behaviors, without counterpart in planar jellium models.

cond-mat.soft↗

arXiv2019

Elasticity in Curved Topographies: Exact Theories and Linear Approximations

Almost all available results in elasticity on curved topographies are obtained within either a small curvature expansion or an empirical covariant generalization that accounts for screening between Gaussian curvature and disclinations. In this paper, we present a formulation of elasticity theory in curved geometries that unifies its underlying geometric and topological content with the theory of defects. The two different linear approximations widely used in the literature are shown to arise as systematic expansions in reference and actual space. Taking the concrete example of a 2D crystal, with and without a central disclination, constrained on a spherical cap, we compare the exact results with different approximations and evaluate their range of validity. We conclude with some general discussion about the universality of non-linear elasticity.

cond-mat.soft↗

arXiv2018

The equilibrium structure of self-assembled protein nano-cages

Understanding how highly symmetric, robust, monodisperse protein cages self-assemble can have major applications in various areas of bio-nanotechnology, such as drug delivery, biomedical imaging and gene therapy. We develop a model to investigate the assembly of protein subunits into the structures with different size and symmetry. Using Monte Carlo simulation, we obtain the global minimum energy structures. Our results suggest that the physical properties of building blocks including the spontaneous curvature, flexibility and bending rigidity of coat proteins are sufficient to predict the size of the assembly products and that the symmetry and shape selectivity of nano-cages can be explained, at least in part, on a thermodynamic basis. The polymorphism of nano-cages observed in vitro assembly experiments are also discussed.

cond-mat.soft↗

arXiv2018

Why large icosahedral viruses need scaffolding proteins: The interplay of Gaussian curvature and disclination interactions

While small single stranded viral shells encapsidate their genome spontaneously, many large viruses, such as the Herpes virus or Infectious Bursal Disease Virus (IBDV), typically require a template, consisting of either scaffolding proteins or inner core. Despite the proliferation of large viruses in nature, the mechanisms by which hundreds or thousands of proteins assemble to form structures with icosahedral order (IO) is completely unknown. Using continuum elasticity theory, we study the growth of large viral shells (capsids) and show that a non-specific template not only selects the radius of the capsid, but leads to the error-free assembly of protein subunits into capsids with universal IO. We prove that as a spherical cap grows, there is a deep potential well at the locations of disclinations that later in the assembly process will become the vertices of an icosahedron. Furthermore, we introduce a minimal model and simulate the assembly of viral shell around a template under non-equilibrium conditions and find a perfect match between the results of continuum elasticity theory and the numerical simulations. Besides explaining available experimental results, we provide a number of new predictions. Implications for other problems in spherical crystals are also discussed.

physics.bio-ph↗

arXiv2018

The effect of RNA stiffness on the self-assembly of virus particles

Under many in vitro conditions, some small viruses spontaneously encapsidate a single stranded (ss) RNA into a protein shell called the capsid. While viral RNAs are found to be compact and highly branched because of long distance base-pairing between nucleotides, recent experiments reveal that in a head-to-head competition between a ssRNA with no secondary or higher order structure and a viral RNA, the capsid proteins preferentially encapsulate the linear polymer! In this paper, we study the impact of genome stiffness on the encapsidation free energy of the complex of RNA and capsid proteins. We show that an increase in effective chain stiffness because of base-pairing could be the reason why under certain conditions linear chains have an advantage over branched chains when it comes to encapsidation efficiency. While branching makes the genome more compact, RNA base-pairing increases the effective Kuhn length of the RNA molecule, which could result in an increase of the free energy of RNA confinement, that is, the work required to encapsidate RNA, and thus less efficient packaging.

physics.bio-ph↗