arXiv subjects

Shuo Huang

Publications and source records attributed to Shuo Huang.

At least 55 records · Page 3Linked to original sources

arXiv2025

ALMA Lensing Cluster Survey: Physical characterization of near-infrared-dark intrinsically faint ALMA sources at z=2-4

We present results from Atacama Large Millimeter/submillimeter Array (ALMA) spectral line-scan observations at 3-mm and 2-mm bands of three near-infrared-dark (NIR-dark) galaxies behind two massive lensing clusters MACS J0417.5-1154 and RXC J0032.1+1808. Each of these three sources is a faint (de-lensed $S_{\text{1.2 mm}}$ $<$ 1 mJy) triply lensed system originally discovered in the ALMA Lensing Cluster Survey. We have successfully detected CO and [C I] emission lines and confirmed that their spectroscopic redshifts are $z=3.652$, 2.391, and 2.985. By utilizing a rich multi-wavelength data set, we find that the NIR-dark galaxies are located on the star formation main sequence in the intrinsic stellar mass range of log ($M_*$/$M_\odot$) = 9.8 - 10.4, which is about one order of magnitude lower than that of typical submillimeter galaxies (SMGs). These NIR-dark galaxies show a variety in gas depletion times and spatial extent of dust emission. One of the three is a normal star-forming galaxy with gas depletion time consistent with a scaling relation, and its infrared surface brightness is an order of magnitude smaller than that of typical SMGs. Since this galaxy has an elongated axis ratio of $\sim 0.17$, we argue that normal star-forming galaxies in an edge-on configuration can be heavily dust-obscured. This implies that existing deep WFC3/F160W surveys may miss a fraction of typical star-forming main-sequence galaxies due to their edge-on orientation.

astro-ph.GA↗

arXiv2024

ssProp: Energy-Efficient Training for Convolutional Neural Networks with Scheduled Sparse Back Propagation

Recently, deep learning has made remarkable strides, especially with generative modeling, such as large language models and probabilistic diffusion models. However, training these models often involves significant computational resources, requiring billions of petaFLOPs. This high resource consumption results in substantial energy usage and a large carbon footprint, raising critical environmental concerns. Back-propagation (BP) is a major source of computational expense during training deep learning models. To advance research on energy-efficient training and allow for sparse learning on any machine and device, we propose a general, energy-efficient convolution module that can be seamlessly integrated into any deep learning architecture. Specifically, we introduce channel-wise sparsity with additional gradient selection schedulers during backward based on the assumption that BP is often dense and inefficient, which can lead to over-fitting and high computational consumption. Our experiments demonstrate that our approach reduces 40\% computations while potentially improving model performance, validated on image classification and generation tasks. This reduction can lead to significant energy savings and a lower carbon footprint during the research and development phases of large-scale AI systems. Additionally, our method mitigates over-fitting in a manner distinct from Dropout, allowing it to be combined with Dropout to further enhance model performance and reduce computational resource usage. Extensive experiments validate that our method generalizes to a variety of datasets and tasks and is compatible with a wide range of deep learning architectures and modules. Code is publicly available at https://github.com/lujiazho/ssProp.

cs.LG↗

arXiv2024

On the origin of transition disk cavities: Pebble-accreting protoplanets vs Super-Jupiters

Protoplanetary disks surrounding young stars are the birth places of planets. Among them, transition disks with inner dust cavities of tens of au are sometimes suggested to host massive companions. Yet, such companions are often not detected. Some transition disks exhibit a large amount of gas inside the dust cavity and relatively high stellar accretion rates, which contradicts typical models of gas-giant-hosting systems. Therefore, we investigate whether a sequence of low-mass planets can produce cavities in the dust disk. We evolve the disks with low-mass accreting embryos in combination with 1D dust transport and 3D pebble accretion, to investigate the reduction of the pebble flux at the embryos' orbits. We vary the planet and disk properties. We find that multiple pebble-accreting planets can efficiently decrease the dust surface density, resulting in dust cavities consistent with transition disks. The number of low-mass planets necessary to sweep up all pebbles decreases with decreasing turbulent strength and is preferred when the dust Stokes number is $10^{-2}-10^{-4}$. Compared to dust rings caused by pressure bumps, those by efficient pebble accretion exhibit more extended outer edges. We also highlight the observational reflections: the transition disks with rings featuring extended outer edges tend to have a large gas content in the dust cavities and rather high stellar accretion rates. We propose that planet-hosting transition disks consist of two groups. In Group A disks, planets have evolved into gas giants, opening deep gaps in the gas disk. Pebbles concentrate in pressure maxima, forming dust rings. In Group B, multiple Neptunes (unable to open deep gas gaps) accrete incoming pebbles, causing the appearance of inner dust cavities. The morphological discrepancy of these rings may aid in distinguishing between the two groups using high-resolution ALMA observations.

astro-ph.EP↗

arXiv2024

Spherical Analysis of Learning Nonlinear Functionals

In recent years, there has been growing interest in the field of functional neural networks. They have been proposed and studied with the aim of approximating continuous functionals defined on sets of functions on Euclidean domains. In this paper, we consider functionals defined on sets of functions on spheres. The approximation ability of deep ReLU neural networks is investigated by novel spherical analysis using an encoder-decoder framework. An encoder comes up first to accommodate the infinite-dimensional nature of the domain of functionals. It utilizes spherical harmonics to help us extract the latent finite-dimensional information of functions, which in turn facilitates in the next step of approximation analysis using fully connected neural networks. Moreover, real-world objects are frequently sampled discretely and are often corrupted by noise. Therefore, encoders with discrete input and those with discrete and random noise input are constructed, respectively. The approximation rates with different encoder structures are provided therein.

cs.LG↗

arXiv2024

PsybORG+: Modeling and Simulation for Detecting Cognitive Biases in Advanced Persistent Threats

Advanced Persistent Threats (APTs) bring significant challenges to cybersecurity due to their sophisticated and stealthy nature. Traditional cybersecurity measures fail to defend against APTs. Cognitive vulnerabilities can significantly influence attackers' decision-making processes, which presents an opportunity for defenders to exploit. This work introduces PsybORG$^+$, a multi-agent cybersecurity simulation environment designed to model APT behaviors influenced by cognitive vulnerabilities. A classification model is built for cognitive vulnerability inference and a simulator is designed for synthetic data generation. Results show that PsybORG$^+$ can effectively model APT attackers with different loss aversion and confirmation bias levels. The classification model has at least a 0.83 accuracy rate in predicting cognitive vulnerabilities.

cs.CR↗

arXiv2024

On the suppression of giant planet formation around low-mass stars in clustered environments

Context: Current exoplanet formation studies tend to overlook the birth environment of stars in clustered environments. The effect of this environment on the planet-formation process, however, is important, especially in the earliest stage. Aims: We investigate the differences in planet populations forming in star-cluster environments through pebble accretion and compare these results with the planet formation around isolated stars. We try to provide potential signatures on the young planetary systems to guide future observation. Methods: We design and present a new planet population synthesis code for clustered environments. The planet formation model is based on pebble accretion and includes migration in the circumstellar disk. The disk's gas and dust are evolved in 1D simulations considering the effects of photo-evaporation of the nearby stars. Results: Planetary systems in a clustered environment are different than those born in isolation; the environmental effects are important for a wide range of observable parameters and the eventual architecture of the planetary systems. Planetary systems born in a clustered environment lack cold Jupiters compared to isolated planetary systems. This effect is more pronounced for low-mass stars ($\lesssim$0.2 $M_\odot$). On the other hand, planetary systems born in clusters show an excess of cold Neptune around these low-mass stars. Conclusions: In future observations, finding an excess of cold Neptunes and a lack of cold Jupiters could be used to constrain the birth environments of these planetary systems. Exploring the dependence of cold Jupiter's intrinsic occurrence rate on stellar mass provides insights into the birth environment of their proto-embryos.

astro-ph.EP↗

arXiv2024

PlacidDreamer: Advancing Harmony in Text-to-3D Generation

Recently, text-to-3D generation has attracted significant attention, resulting in notable performance enhancements. Previous methods utilize end-to-end 3D generation models to initialize 3D Gaussians, multi-view diffusion models to enforce multi-view consistency, and text-to-image diffusion models to refine details with score distillation algorithms. However, these methods exhibit two limitations. Firstly, they encounter conflicts in generation directions since different models aim to produce diverse 3D assets. Secondly, the issue of over-saturation in score distillation has not been thoroughly investigated and solved. To address these limitations, we propose PlacidDreamer, a text-to-3D framework that harmonizes initialization, multi-view generation, and text-conditioned generation with a single multi-view diffusion model, while simultaneously employing a novel score distillation algorithm to achieve balanced saturation. To unify the generation direction, we introduce the Latent-Plane module, a training-friendly plug-in extension that enables multi-view diffusion models to provide fast geometry reconstruction for initialization and enhanced multi-view images to personalize the text-to-image diffusion model. To address the over-saturation problem, we propose to view score distillation as a multi-objective optimization problem and introduce the Balanced Score Distillation algorithm, which offers a Pareto Optimal solution that achieves both rich details and balanced saturation. Extensive experiments validate the outstanding capabilities of our PlacidDreamer. The code is available at \url{https://github.com/HansenHuang0823/PlacidDreamer}.

cs.CV↗

arXiv2024

Diffusion Model-based FOD Restoration from High Distortion in dMRI

Fiber orientation distributions (FODs) is a popular model to represent the diffusion MRI (dMRI) data. However, imaging artifacts such as susceptibility-induced distortion in dMRI can cause signal loss and lead to the corrupted reconstruction of FODs, which prohibits successful fiber tracking and connectivity analysis in affected brain regions such as the brain stem. Generative models, such as the diffusion models, have been successfully applied in various image restoration tasks. However, their application on FOD images poses unique challenges since FODs are 4-dimensional data represented by spherical harmonics (SPHARM) with the 4-th dimension exhibiting order-related dependency. In this paper, we propose a novel diffusion model for FOD restoration that can recover the signal loss caused by distortion artifacts. We use volume-order encoding to enhance the ability of the diffusion model to generate individual FOD volumes at all SPHARM orders. Moreover, we add cross-attention features extracted across all SPHARM orders in generating every individual FOD volume to capture the order-related dependency across FOD volumes. We also condition the diffusion model with low-distortion FODs surrounding high-distortion areas to maintain the geometric coherence of the generated FODs. We trained and tested our model using data from the UK Biobank (n = 1315). On a test set with ground truth (n = 43), we demonstrate the high accuracy of the generated FODs in terms of root mean square errors of FOD volumes and angular errors of FOD peaks. We also apply our method to a test set with large distortion in the brain stem area (n = 1172) and demonstrate the efficacy of our method in restoring the FOD integrity and, hence, greatly improving tractography performance in affected brain regions.

eess.IV↗

arXiv2024

A monotone scheme for G-equations with application to the explicit convergence rate of robust central limit theorem

We propose a monotone approximation scheme for a class of fully nonlinear PDEs called G-equations. Such equations arise often in the characterization of G-distributed random variables in a sublinear expectation space. The proposed scheme is constructed recursively based on a piecewise constant approximation of the viscosity solution to the G-equation. We establish the convergence of the scheme and determine the convergence rate with an explicit error bound, using the comparison principles for both the scheme and the equation together with a mollification procedure. The first application is obtaining the convergence rate of Peng's robust central limit theorem with an explicit bound of Berry-Esseen type. The second application is an approximation scheme with its convergence rate for the Black-Scholes-Barenblatt equation.

math.PR↗

arXiv2023

J0107a: A Barred Spiral Dusty Star-forming Galaxy at $z=2.467$

Dusty Star-Forming Galaxies (DSFGs) are amongst the most massive and active star-forming galaxies during the cosmic noon. Theoretical studies have proposed various formation mechanisms of DSFGs, including major merger-driven starbursts and secular star-forming disks. Here, we report J0107a, a bright ($\sim8$ mJy at observed-frame 888 $μ$m) DSFG at $z=2.467$ that appears to be a gas-rich massive disk and might be an extreme case of the secular disk scenario. J0107a has a stellar mass $M_\star\sim5\times10^{11}M_\odot$, molecular gas mass $M_\mathrm{mol}\sim(1\textendash6)\times10^{11}M_\odot$, and a star formation rate (SFR) of $\sim500M_\odot$ yr$^{-1}$. J0107a does not have a gas-rich companion. The rest-frame 1.28 $μ$m JWST NIRCam image of J0107a shows a grand-design spiral with a prominent stellar bar extending $\sim15$ kpc. ALMA band 7 continuum map reveals that the dust emission originates from both the central starburst and the stellar bar. 3D disk modeling of the CO(4-3) emission line indicates a dynamically cold disk with rotation-to-dispersion ratio $V_\mathrm{max}/σ\sim8$. The results suggest a bright DSFG may have a non-merger origin, and its vigorous star formation may be triggered by bar and/or rapid gas inflow.

astro-ph.GA↗

arXiv2023

AvatarFusion: Zero-shot Generation of Clothing-Decoupled 3D Avatars Using 2D Diffusion

Large-scale pre-trained vision-language models allow for the zero-shot text-based generation of 3D avatars. The previous state-of-the-art method utilized CLIP to supervise neural implicit models that reconstructed a human body mesh. However, this approach has two limitations. Firstly, the lack of avatar-specific models can cause facial distortion and unrealistic clothing in the generated avatars. Secondly, CLIP only provides optimization direction for the overall appearance, resulting in less impressive results. To address these limitations, we propose AvatarFusion, the first framework to use a latent diffusion model to provide pixel-level guidance for generating human-realistic avatars while simultaneously segmenting clothing from the avatar's body. AvatarFusion includes the first clothing-decoupled neural implicit avatar model that employs a novel Dual Volume Rendering strategy to render the decoupled skin and clothing sub-models in one space. We also introduce a novel optimization method, called Pixel-Semantics Difference-Sampling (PS-DS), which semantically separates the generation of body and clothes, and generates a variety of clothing styles. Moreover, we establish the first benchmark for zero-shot text-to-avatar generation. Our experimental results demonstrate that our framework outperforms previous approaches, with significant improvements observed in all metrics. Additionally, since our model is clothing-decoupled, we can exchange the clothes of avatars. Code are available on our project page https://hansenhuang0823.github.io/AvatarFusion.

cs.CV↗

arXiv2023

Generating Redstone Style Cities in Minecraft

Procedurally generating cities in Minecraft provides players more diverse scenarios and could help understand and improve the design of cities in other digital worlds and the real world. This paper presents a city generator that was submitted as an entry to the 2023 Edition of Minecraft Settlement Generation Competition for Minecraft. The generation procedure is composed of six main steps, namely vegetation clearing, terrain reshaping, building layout generation, route planning, streetlight placement, and wall construction. Three algorithms, including a heuristic-based algorithm, an evolving layout algorithm, and a random one are applied to generate the building layout, thus determining where to place different redstone style buildings, and tested by generating cities on random maps in limited time. Experimental results show that the heuristic-based algorithm is capable of finding an acceptable building layout faster for flat maps, while the evolving layout algorithm performs better in evolving layout for rugged maps. A user study is conducted to compare our generator with outstanding entries of the competition's 2022 edition using the competition's evaluation criteria and shows that our generator performs well in the adaptation and functionality criteria

cs.AI↗

arXiv2023

Characterizing CO Emitters in the SSA22-AzTEC26 Field

We report the physical characterization of four CO emitters detected near the bright submillimeter galaxy (SMG) SSA22-AzTEC26. We analyze the data from Atacama Large Millimeter/submillileter Array band 3, 4, and 7 observations of the SSA22-AzTEC26 field. In addition to the targeted SMG, we detect four line emitters with a signal-to-noise ratio $>5.2$ in the cube smoothed with 300 km s$^{-1}$ FWHM Gaussian filter. All four sources have NIR counterparts within 1$\arcsec$. We perform UV-to-FIR spectral energy distribution modeling to derive the photometric redshifts and physical properties. Based on the photometric redshifts, we reveal that two of them are CO(2-1) at redshifts of 1.113 and 1.146 and one is CO(3-2) at $z=2.124$. The three sources are massive galaxies with a stellar mass $\gtrsim10^{10.5}M_\odot$, but have different levels of star formation. Two lie within the scatter of the main sequence (MS) of star-forming galaxies at $z\sim1-2$, and the most massive galaxy lies significantly below the MS. However, all three sources have a gas fraction within the scatter of the MS scaling relation. This shows that a blind CO line search can detect massive galaxies with low specific star formation rates that still host large gas reservoirs and that it also complements targeted surveys, suggesting later gas acquisition and the need for other mechanisms in addition to gas consumption to suppress star formation.

astro-ph.GA↗

arXiv2023

Shuffled Autoregression For Motion Interpolation

This work aims to provide a deep-learning solution for the motion interpolation task. Previous studies solve it with geometric weight functions. Some other works propose neural networks for different problem settings with consecutive pose sequences as input. However, motion interpolation is a more complex problem that takes isolated poses (e.g., only one start pose and one end pose) as input. When applied to motion interpolation, these deep learning methods have limited performance since they do not leverage the flexible dependencies between interpolation frames as the original geometric formulas do. To realize this interpolation characteristic, we propose a novel framework, referred to as \emph{Shuffled AutoRegression}, which expands the autoregression to generate in arbitrary (shuffled) order and models any inter-frame dependencies as a directed acyclic graph. We further propose an approach to constructing a particular kind of dependency graph, with three stages assembled into an end-to-end spatial-temporal motion Transformer. Experimental results on one of the current largest datasets show that our model generates vivid and coherent motions from only one start frame to one end frame and outperforms competing methods by a large margin. The proposed model is also extensible to multiple keyframes' motion interpolation tasks and other areas' interpolation.

cs.CV↗

arXiv2023

Energy-Latency Attacks to On-Device Neural Networks via Sponge Poisoning

In recent years, on-device deep learning has gained attention as a means of developing affordable deep learning applications for mobile devices. However, on-device models are constrained by limited energy and computation resources. In the mean time, a poisoning attack known as sponge poisoning has been developed.This attack involves feeding the model with poisoned examples to increase the energy consumption during inference. As previous work is focusing on server hardware accelerators, in this work, we extend the sponge poisoning attack to an on-device scenario to evaluate the vulnerability of mobile device processors. We present an on-device sponge poisoning attack pipeline to simulate the streaming and consistent inference scenario to bridge the knowledge gap in the on-device setting. Our exclusive experimental analysis with processors and on-device networks shows that sponge poisoning attacks can effectively pollute the modern processor with its built-in accelerator. We analyze the impact of different factors in the sponge poisoning algorithm and highlight the need for improved defense mechanisms to prevent such attacks on on-device deep learning applications.

cs.CR↗

arXiv2023

Beyond the Model: Data Pre-processing Attack to Deep Learning Models in Android Apps

The increasing popularity of deep learning (DL) models and the advantages of computing, including low latency and bandwidth savings on smartphones, have led to the emergence of intelligent mobile applications, also known as DL apps, in recent years. However, this technological development has also given rise to several security concerns, including adversarial examples, model stealing, and data poisoning issues. Existing works on attacks and countermeasures for on-device DL models have primarily focused on the models themselves. However, scant attention has been paid to the impact of data processing disturbance on the model inference. This knowledge disparity highlights the need for additional research to fully comprehend and address security issues related to data processing for on-device models. In this paper, we introduce a data processing-based attacks against real-world DL apps. In particular, our attack could influence the performance and latency of the model without affecting the operation of a DL app. To demonstrate the effectiveness of our attack, we carry out an empirical study on 517 real-world DL apps collected from Google Play. Among 320 apps utilizing MLkit, we find that 81.56\% of them can be successfully attacked. The results emphasize the importance of DL app developers being aware of and taking actions to secure on-device models from the perspective of data processing.

cs.CR↗

arXiv2023

When, where, and how many planets end up in first-order resonances?

The theory of Type~I migration has been widely used in many studies. Transiting multi-planet systems offer us the opportunity to examine the consistency between observation and theory, especially for those systems harbouring planets in Mean Motion Resonance (MMR). The displacement these resonant pairs show from exact commensurability provides us with information on their migration and eccentricity-damping histories. Here, we adopt a probabilistic approach, characterized by two distributions -- appropriate for either the resonant or non-resonant planets -- to fit the observed planet period ratio distribution. With the Markov chain Monte Carlo (MCMC) method, we find that ${\approx}15\%$ of exoplanets are in first order ($j+1{:}j$) MMRs, the ratio of eccentricity-to-semi-major axis damping is too high to allow overstable librations and that the results are by-and-large consistent with Type-I migration theory. In addition, our modeling finds that a small fraction of resonant pairs is captured into resonance during migration, implying late planet formation (gas-poor). Most of the resonant pairs park themselves at the migration barrier, indicating early planet formation (gas-rich). Furthermore, after improving the criterion on two-body resonant trapping, we obtain an upper limit of the disc surface density at the time the planets are locked in resonance.

astro-ph.EP↗

arXiv2022

Site-specific atomic substitution in a giant magnetocaloric Fe$_2$P-type system

Giant magnetocaloric (GMC) materials constitute a requirement for near room temperature magnetic refrigeration. (Fe,Mn)$_2$(P,Si) is a GMC compound with strong magnetoelastic coupling. The main hindrance towards application of this material is a comparably large temperature hysteresis, which can be reduced by metal site substitution with a nonmagnetic element. However, the (Fe,Mn)$_2$(P,Si) compound has two equally populated metal sites, the tetrahedrally coordinated $3f$ and the pyramidally coordinated $3g$ sites. The magnetic and magnetocaloric properties of such compounds are highly sensitive to the site specific occupancy of the magnetic atoms. Here we have attempted to study separately the effect of $3f$ and $3g$ site substitution with equal amounts of vanadium. Using formation energy calculations, the site preference of vanadium and its influence on the magnetic phase formation are described. A large difference in the isothermal entropy change (as high as 44\%) with substitution in the $3f$ and $3g$ sites is observed. The role of the lattice parameter change with temperature and the strength of the magnetoelastic coupling on the magnetic properties are highlighted.

cond-mat.mtrl-sci↗