arXiv subjects

Siyuan Zhang

Publications and source records attributed to Siyuan Zhang.

At least 73 records · Page 4Linked to original sources

arXiv2025

Using MRNet to Predict Lunar Rock Categories Detected by Chang'e 5 Probe

China's Chang'e 5 mission has been a remarkable success, with the chang'e 5 lander traveling on the Oceanus Procellarum to collect images of the lunar surface. Over the past half century, people have brought back some lunar rock samples, but its quantity does not meet the need for research. Under current circumstances, people still mainly rely on the analysis of rocks on the lunar surface through the detection of lunar rover. The Oceanus Procellarum, chosen by Chang'e 5 mission, contains various kind of rock species. Therefore, we first applied to the National Astronomical Observatories of the China under the Chinese Academy of Sciences for the Navigation and Terrain Camera (NaTeCam) of the lunar surface image, and established a lunar surface rock image data set CE5ROCK. The data set contains 100 images, which randomly divided into training, validation and test set. Experimental results show that the identification accuracy testing on convolutional neural network (CNN) models like AlexNet or MobileNet is about to 40.0%. In order to make full use of the global information in Moon images, this paper proposes the MRNet (MoonRockNet) network architecture. The encoding structure of the network uses VGG16 for feature extraction, and the decoding part adds dilated convolution and commonly used U-Net structure on the original VGG16 decoding structure, which is more conducive to identify more refined but more sparsely distributed types of lunar rocks. We have conducted extensive experiments on the established CE5ROCK data set, and the experimental results show that MRNet can achieve more accurate rock type identification, and outperform other existing mainstream algorithms in the identification performance.

cs.CV↗

arXiv2024

Parametric $ρ$-Norm Scaling Calibration

Output uncertainty indicates whether the probabilistic properties reflect objective characteristics of the model output. Unlike most loss functions and metrics in machine learning, uncertainty pertains to individual samples, but validating it on individual samples is unfeasible. When validated collectively, it cannot fully represent individual sample properties, posing a challenge in calibrating model confidence in a limited data set. Hence, it is crucial to consider confidence calibration characteristics. To counter the adverse effects of the gradual amplification of the classifier output amplitude in supervised learning, we introduce a post-processing parametric calibration method, $ρ$-Norm Scaling, which expands the calibrator expression and mitigates overconfidence due to excessive amplitude while preserving accuracy. Moreover, bin-level objective-based calibrator optimization often results in the loss of significant instance-level information. Therefore, we include probability distribution regularization, which incorporates specific priori information that the instance-level uncertainty distribution after calibration should resemble the distribution before calibration. Experimental results demonstrate the substantial enhancement in the post-processing calibrator for uncertainty calibration with our proposed method.

cs.LG↗

arXiv2024

Color-Kinematics Duality and Dual Conformal Symmetry for A Four-loop Form Factor in N=4 SYM

We obtain the full-color four-loop three-point form factor of the stress-tensor supermultiplet in N=4 SYM, based on the color-kinematics (CK) duality and generalized unitarity method. The CK-dual solution, while manifesting all dual Jacobi relations and satisfying the minimal power-counting of loop momenta, lies in a 133-dimensional solution space. We also show that the planar form factor integrand satisfies precisely a directional dual conformal symmetry in the lightlike limit of the operator momentum, which is supported by explicit four-loop calculations.

hep-th↗

arXiv2024

Evaluating Fairness in Black-box Algorithmic Markets: A Case Study of Ride Sharing in Chicago

This study examines fairness within the rideshare industry, focusing on both drivers' wages and riders' trip fares. Through quantitative analysis, we found that drivers' hourly wages are significantly influenced by factors such as race/ethnicity, health insurance status, tenure to the platform, and working hours. Despite platforms' policies not intentionally embedding biases, disparities persist based on these characteristics. For ride fares, we propose a method to audit the pricing policy of a proprietary algorithm by replicating it; we conduct a hypothesis test to determine if the predicted rideshare fare is greater than the taxi fare, taking into account the approximation error in the replicated model. Challenges in accessing data and transparency hinder our ability to isolate discrimination from other factors, underscoring the need for collaboration with rideshare platforms and drivers to enhance fairness in algorithmic wage determination and pricing.

cs.HC↗

arXiv2024

MER 2024: Semi-Supervised Learning, Noise Robustness, and Open-Vocabulary Multimodal Emotion Recognition

Multimodal emotion recognition is an important research topic in artificial intelligence. Over the past few decades, researchers have made remarkable progress by increasing the dataset size and building more effective algorithms. However, due to problems such as complex environments and inaccurate annotations, current systems are hard to meet the demands of practical applications. Therefore, we organize the MER series of competitions to promote the development of this field. Last year, we launched MER2023, focusing on three interesting topics: multi-label learning, noise robustness, and semi-supervised learning. In this year's MER2024, besides expanding the dataset size, we further introduce a new track around open-vocabulary emotion recognition. The main purpose of this track is that existing datasets usually fix the label space and use majority voting to enhance the annotator consistency. However, this process may lead to inaccurate annotations, such as ignoring non-majority or non-candidate labels. In this track, we encourage participants to generate any number of labels in any category, aiming to describe emotional states as accurately as possible. Our baseline code relies on MERTools and is available at: https://github.com/zeroQiaoba/MERTools/tree/master/MER2024.

cs.LG↗

arXiv2024

Grain boundaries control lithiation of solid solution substrates in lithium metal batteries

The development of sustainable transportation and communication systems requires an increase in both energy density and capacity retention of Li-batteries. Using substrates forming a solid solution with body centered cubic Li enhances the cycle stability of anode-less batteries. However, it remains unclear how the substrate microstructure affects the lithiation behavior. Here, we deploy a correlative, near-atomic scale probing approach through combined ion- and electron-microscopy to examine the distribution of Li in Li-Ag diffusion couples as model system. We reveal that Li regions with over 93.8% at.% nucleate within Ag at random high angle grain boundaries, whereas grain interiors are not lithiated. We evidence the role of kinetics and mechanical constraint from the microstructure over equilibrium thermodynamics in dictating the lithiation process. The findings suggest that grain size and grain boundary character are critical to enhance the electrochemical performance of interlayers/electrodes, particularly for improving lithiation kinetics and hence reducing dendrite formation.

cond-mat.mtrl-sci↗

arXiv2024

Explainable Multimodal Emotion Recognition

Multimodal emotion recognition is an important research topic in artificial intelligence, whose main goal is to integrate multimodal clues to identify human emotional states. Current works generally assume accurate labels for benchmark datasets and focus on developing more effective architectures. However, emotion annotation relies on subjective judgment. To obtain more reliable labels, existing datasets usually restrict the label space to some basic categories, then hire plenty of annotators and use majority voting to select the most likely label. However, this process may result in some correct but non-candidate or non-majority labels being ignored. To ensure reliability without ignoring subtle emotions, we propose a new task called ``Explainable Multimodal Emotion Recognition (EMER)''. Unlike traditional emotion recognition, EMER takes a step further by providing explanations for these predictions. Through this task, we can extract relatively reliable labels since each label has a certain basis. Meanwhile, we borrow large language models (LLMs) to disambiguate unimodal clues and generate more complete multimodal explanations. From them, we can extract richer emotions in an open-vocabulary manner. This paper presents our initial attempt at this task, including introducing a new dataset, establishing baselines, and defining evaluation metrics. In addition, EMER can serve as a benchmark task to evaluate the audio-video-text understanding performance of multimodal LLMs.

cs.MM↗

arXiv2024

Operando Insights on the Degradation Mechanisms of Rhenium-doped and Undoped Molybdenum Disulfide Nanocatalysts for Electrolyzer Applications

MoS2 nanostructures are promising catalysts for proton-exchange-membrane (PEM) electrolyzers to replace expensive noble metals. Their broadscale application demands high activity for the hydrogen evolution reaction (HER) as well as robust durability. Doping is commonly applied to enhance the HER activity of MoS2-based nanocatalysts, but the effect of dopants in the electrochemical and structural stability is yet to be discussed. Herein, we correlate operando electrochemical measurements to the structural evolution of the materials down to the nanometric scale by identical location electron microscopy and spectroscopy. The range of stable operation for MoS2 nanocatalysts with and without rhenium doping is experimentally defined. The responsible degradation mechanisms at first electrolyte contact, open circuit stabilization and HER conditions are experimentally identified and confirmed with the calculated Pourbaix diagram of Re-doped MoS2. Doping MoS2-based nanocatalysts is validated as a promising strategy for the continuous improvement of high performance and durable PEM electrolyzers.

physics.app-ph↗

arXiv2024

Exploring the Transferability of Visual Prompting for Multimodal Large Language Models

Although Multimodal Large Language Models (MLLMs) have demonstrated promising versatile capabilities, their performance is still inferior to specialized models on downstream tasks, which makes adaptation necessary to enhance their utility. However, fine-tuning methods require independent training for every model, leading to huge computation and memory overheads. In this paper, we propose a novel setting where we aim to improve the performance of diverse MLLMs with a group of shared parameters optimized for a downstream task. To achieve this, we propose Transferable Visual Prompting (TVP), a simple and effective approach to generate visual prompts that can transfer to different models and improve their performance on downstream tasks after trained on only one model. We introduce two strategies to address the issue of cross-model feature corruption of existing visual prompting methods and enhance the transferability of the learned prompts, including 1) Feature Consistency Alignment: which imposes constraints to the prompted feature changes to maintain task-agnostic knowledge; 2) Task Semantics Enrichment: which encourages the prompted images to contain richer task-specific semantics with language guidance. We validate the effectiveness of TVP through extensive experiments with 6 modern MLLMs on a wide variety of tasks ranging from object recognition and counting to multimodal reasoning and hallucination correction.

cs.CV↗

arXiv2024

Fabrication and extreme micromechanics of additive metal microarchitectures

The mechanical performance of metallic metamaterials with 3-dimensional solid frames is typically a combination of the geometrical effect ("architecture") and the characteristic size effects of the base material ("microstructure"). In this study, for the first time, the temperature- and rate-dependent mechanical response of copper microlattices has been investigated. The microlattices were fabricated via a localized electrodeposition in liquid (LEL) process which enables high-precision additive manufacturing of metal at the micro-scale. The metal microlattices possess a unique microstructure with micron sized grains that are rich with randomly oriented growth twins and near-ideal nodal connectivity. Importantly, copper microlattices exhibited unique temperature (-150 and 25 degree C) and strain rate (0.001~100 s-1) dependent deformation behavior during in situ micromechanical testing. Systematic compression tests of fully dense copper micropillars, equivalent in diameter and length to the struts of the microlattice at comparable extreme loading conditions, allow us to investigate the intrinsic deformation mechanism of copper. Combined with the post-mortem microstructural analysis, substantial shifts in deformation mechanisms depending on the temperature and strain rate were revealed. On the one hand, at room temperature (25 degree C), dislocation slip based plastic deformation occurs and leads to a localized deformation of the micropillars. On the other hand, at cryogenic temperature (-150 degree C), mechanical twinning occurs and leads to relatively homogeneous deformation of the micropillars. Based on the intrinsic deformation mechanisms of copper, the temperature and strain rate dependent deformation behavior of microlattices could be explained.

cond-mat.mtrl-sci↗

arXiv2024

Multimodal Fusion with Pre-Trained Model Features in Affective Behaviour Analysis In-the-wild

Multimodal fusion is a significant method for most multimodal tasks. With the recent surge in the number of large pre-trained models, combining both multimodal fusion methods and pre-trained model features can achieve outstanding performance in many multimodal tasks. In this paper, we present our approach, which leverages both advantages for addressing the task of Expression (Expr) Recognition and Valence-Arousal (VA) Estimation. We evaluate the Aff-Wild2 database using pre-trained models, then extract the final hidden layers of the models as features. Following preprocessing and interpolation or convolution to align the extracted features, different models are employed for modal fusion. Our code is available at GitHub - FulgenceWen/ABAW6th.

cs.CV↗

arXiv2024

Sub-GeV events energy reconstruction with 3-inch PMTs in JUNO

A 20-kiloton liquid scintillator detector is designed in the Jiangmen Underground Neutrino Observatory (JUNO) for multiple physics purposes, including the determination of the neutrino mass ordering through reactor neutrinos, as well as measuring supernova neutrinos, solar neutrinos, and atmosphere neutrinos to explore different physics topics. Efficient reconstruction algorithms are needed to achieve these physics goals in a wide energy range from MeV to GeV. In this paper, we present a novel method for reconstructing the energy of events using hit information from 25600 3-inch photomultiplier tubes (PMTs) and the OCCUPANCY method. Our algorithm exhibits good performance in accurate energy reconstruction, validated with electron Monte Carlo samples spanning kinetic energies from 10 MeV to 1 GeV.

physics.ins-det↗

arXiv2023

GraphScope Flex: LEGO-like Graph Computing Stack

Graph computing has become increasingly crucial in processing large-scale graph data, with numerous systems developed for this purpose. Two years ago, we introduced GraphScope as a system addressing a wide array of graph computing needs, including graph traversal, analytics, and learning in one system. Since its inception, GraphScope has achieved significant technological advancements and gained widespread adoption across various industries. However, one key lesson from this journey has been understanding the limitations of a "one-size-fits-all" approach, especially when dealing with the diversity of programming interfaces, applications, and data storage formats in graph computing. In response to these challenges, we present GraphScope Flex, the next iteration of GraphScope. GraphScope Flex is designed to be both resource-efficient and cost-effective, while also providing flexibility and user-friendliness through its LEGO-like modularity. This paper explores the architectural innovations and fundamental design principles of GraphScope Flex, all of which are direct outcomes of the lessons learned during our ongoing development process. We validate the adaptability and efficiency of GraphScope Flex with extensive evaluations on synthetic and real-world datasets. The results show that GraphScope Flex achieves 2.4X throughput and up to 55.7X speedup over other systems on the LDBC Social Network and Graphalytics benchmarks, respectively. Furthermore, GraphScope Flex accomplishes up to a 2,400X performance gain in real-world applications, demonstrating its proficiency across a wide range of graph computing scenarios with increased effectiveness.

cs.DC↗

arXiv2023

Probability-Dependent Gradient Decay in Large Margin Softmax

In the past few years, Softmax has become a common component in neural network frameworks. In this paper, a gradient decay hyperparameter is introduced in Softmax to control the probability-dependent gradient decay rate during training. By following the theoretical analysis and empirical results of a variety of model architectures trained on MNIST, CIFAR-10/100 and SVHN, we find that the generalization performance depends significantly on the gradient decay rate as the confidence probability rises, i.e., the gradient decreases convexly or concavely as the sample probability increases. Moreover, optimization with the small gradient decay shows a similar curriculum learning sequence where hard samples are in the spotlight only after easy samples are convinced sufficiently, and well-separated samples gain a higher gradient to reduce intra-class distance. Based on the analysis results, we can provide evidence that the large margin Softmax will affect the local Lipschitz constraint of the loss function by regulating the probability-dependent gradient decay rate. This paper provides a new perspective and understanding of the relationship among concepts of large margin Softmax, local Lipschitz constraint and curriculum learning by analyzing the gradient decay rate. Besides, we propose a warm-up strategy to dynamically adjust Softmax loss in training, where the gradient decay rate increases from over-small to speed up the convergence rate.

stat.ML↗

arXiv2023

Constructing phase diagrams for defects by correlated atomic-scale characterization

Phase transformations and crystallographic defects are two essential tools to drive innovations in materials. Bulk materials design via tuning chemical compositions has been systematized using phase diagrams. We show here that the same thermodynamic concept can be applied to understand the chemistry at defects. We present a combined experimental and modelling approach to scope and build phase diagrams for defects. The discovery was enabled by triggering phase transformations of individual defects through local alloying, and sequentially imaging the structural and chemical changes using atomic-resolution scanning transmission electron microscopy. By observing atomic-scale phase transformations of a Mg grain boundary through Ga alloying, we exemplified the method to construct a grain boundary phase diagram using ab initio simulations and thermodynamic principles. The methodology enables a systematic development of defect phase diagrams to propel a new paradigm for materials design utilizing chemical complexity and phase transformations at defects.

cond-mat.mtrl-sci↗

arXiv2023

Chemical and structural characterization of the native oxide scale on a Mg-based alloy

In this study, the structure and composition of the native oxide forming on the basal plane (0001) of Mg-2Al-0.1Ca is investigated by a correlative approach, combining scanning transmission electron microscopy (STEM) and atom probe tomography (APT). Atom probe specimens were prepared conventionally in a Ga focused ion beam (FIB) as well as a Xe plasma FIB in a cryogenic setup and subsequently cleaned in the atom probe to remove surface contamination before oxidation. While thermal energy input from the laser and longer atmospheric exposure time increased the measured hydrogen content in the specimen's apex region, cryo preparation revealed, that the hydrogen uptake in magnesium is independent of the employment of conventional or cryogenic FIB preparation. TEM measurements demonstrated the growth of a (111) MgO oxide layer with 3-4 nm thickness on the basal (0001) plane of the Mg atom probe specimen. APT data further revealed the formation of an aluminum-rich region between bulk Mg and the native oxide. The aluminum enrichment of up to ~20 at.% at the interface is consistent with an inward growth of the oxide scale.

cond-mat.mtrl-sci↗

arXiv2023

Machine learning-enabled tomographic imaging of chemical short-range atomic ordering

In solids, chemical short-range order (CSRO) refers to the self-organisation of atoms of certain species occupying specific crystal sites. CSRO is increasingly being envisaged as a lever to tailor the mechanical and functional properties of materials. Yet quantitative relationships between properties and the morphology, number density, and atomic configurations of CSRO domains remain elusive. Herein, we showcase how machine learning-enhanced atom probe tomography (APT) can mine the near-atomically resolved APT data and jointly exploit the technique's high elemental sensitivity to provide a 3D quantitative analysis of CSRO in a CoCrNi medium-entropy alloy. We reveal multiple CSRO configurations, with their formation supported by state-of-the-art Monte-Carlo simulations. Quantitative analysis of these CSROs allows us to establish relationships between processing parameters and physical properties. The unambiguous characterization of CSRO will help refine strategies for designing advanced materials by manipulating atomic-scale architectures.

cond-mat.mtrl-sci↗

arXiv2023

Contribution of the oxygen reduction reaction to the electrochemical cathodic partial reaction for Mg-Al-Ca solid solutions

The electrochemical corrosion rate of Magnesium (Mg) and Mg alloys depends on the stability of the formed surface layer. Based on the Mg substrate, the oxide structure comprises a dense MgO/Mg(OH)2 mixture underneath a porous plate-like Mg(OH)2 layer. While the kinetics of the anodic partial reaction has been mainly attributed to MgO, recent studies showed an effect of the Mg(OH)2 layer thickness on the cathodic partial reaction. A thinner Mg(OH)2 layer has been associated with a higher kinetics of the oxygen reduction rate. In the present study, the proposed mechanism has been further investigated via in situ respirometric measurements with Mg-Al-Ca solid solution in electrolytes with different pH values (pH = 8-13). The results indicate an additional effect based on the structure of the surface layer in the passive state of Mg corrosion. Furthermore, two different Al enriched interlayers at the Mg/MgO- and MgO/Mg(OH)2 interfaces were observed and discussed in terms of their thermodynamic stability under alkaline immersion conditions.

physics.chem-ph↗