arXiv subjects

Yiming Ding

Publications and source records attributed to Yiming Ding.

At least 19 recordsLinked to original sources

arXiv2026

Flow-polarity decoupling and universal mobility enhancement in dense bacterial active fluids with mesoscale order

Active fluids consisting of living cells or synthetic microswimmers display rich emergent behavior and nonequilibrium mechanical properties, which not only shed light on various biological processes but also inform the engineering of autonomous fluidics and self-driven materials. The individual behavior of microswimmers and their interaction with self-generated mesoscale solvent flows underlie the emergent properties of active fluids. Here we studied the microscopic dynamics in dense 3D bacterial active fluids by simultaneous imaging of cell body, flagella, and flow field. A surprising finding is that the polarity of cells was randomly distributed in mesoscale flow regimes, and yet the system displays mesoscale order in the self-generated solvent flows. Despite the apparent flow-polarity decoupling, the motion of cells relative to local solvent flows predominantly navigated upstream, with the self-advection speed universally enhanced by a flow-controlled constant. Numerical modeling with full hydrodynamic interactions reveals that the observed flow-polarity decoupling arises from the breakdown of the commonly held force-dipole assumption for anisotropic microswimmers: in the presence of flow gradient and near-field hydrodynamic interactions, the direction of total active forcing exerted by a swimming bacterium to the surrounding fluid no longer aligns with its polarity. The simulations suggest that near-field interactions serve as a new type of emergent, configuration-dependent active forcing, which profoundly impact self-organization and transport in dense bacterial suspensions. Taken together, our work establishes fundamental knowledge for faithfully understanding the collective behavior of dense polar active fluids.

physics.bio-ph↗

arXiv2026

Deterministic fabrication of large-area, high-crystallinity oxide moire superlattices

Oxide twistronics extends moire engineering beyond van der Waals materials, offering a promising platform for accessing emergent interfacial phenomena arising from the strong coupling of lattice, charge, and orbital degrees of freedom in complex oxides. However, deterministic fabrication of high-crystallinity oxide moire superlattices over large lateral dimensions remains challenging due to the three-dimensional bonding network of oxides. Here, we demonstrate a scalable, generalized fabrication strategy that enables the formation of high-crystallinity oxide moire superlattices with clean, chemically bonded interfaces and precisely controlled twist angles down to nominal values of 0.1 degree, achieving sub-degree twist-angle accuracy across large contiguous lateral dimensions approaching the millimeter scale. Using NaNbO3 as a model system, we show that the resulting interlayer coupling drives pronounced structural reconstruction that modifies both the phase structure and ferroelectric domain configuration. Synchrotron-based X-ray 3D reciprocal space mapping reveals the emergence of a single-phase state in twisted bilayers, in contrast to the mixed-phase structure observed in single-layer membranes prior to twist assembly. The structural signatures are further consistent with gradual lattice rotation distributed along the thickness direction that may accommodate interfacial shear strain, distinct from reconstruction observed in van der Waals moire systems, which primarily occurs through in-plane stacking rearrangement. This collective lattice response is correlated with twist-dependent nanoscale electromechanical modulations observed by piezoresponse force microscopy. These results establish a scalable materials platform for oxide twistronics and open new pathways towards integrating twist-engineered complex oxides into practical, macroscale device architectures.

cond-mat.mtrl-sci↗

arXiv2026

Retrieving Any Relevant Moments: Benchmark and Models for Generalized Moment Retrieval

Video Moment Retrieval (VMR) aims to localize temporal segments in videos that correspond to a natural language query, but typically assumes only a single matching moment for each query. This assumption does not always hold in real-world scenarios, where queries may correspond to multiple or no moments. Thus, we formulate Generalized Moment Retrieval (GMR), a unified setting that requires retrieving the complete set of relevant moments or predicting an empty set. To enable systematic study of GMR, we introduce Soccer-GMR, a large-scale benchmark built on challenging soccer videos that reflect general GMR scenarios, with realistic negative and positive queries. The benchmark is constructed via a duration-flexible semi-automated pipeline with human verification, enabling scalable data generation while maintaining high annotation quality. We further design a unified evaluation protocol with complementary metrics tailored for null-set rejection, positive-query localization, and end-to-end GMR performance. Finally, we establish strong baselines across two modeling paradigms: a lightweight plug-and-play GMR adapter for discriminative VMR models, and a GMR-tailored GRPO reward for fine-tuning multimodal large language models (MLLMs). Extensive experiments show consistent gains across all metrics and expose key limitations of current methods, positioning GMR as a more realistic and challenging benchmark for video-language understanding.

cs.CV↗

arXiv2025

Ultrafast ultrasound coded vector Doppler imaging of blood flow velocity and resistivity

Dynamic and precise measurement of cerebral blood flow velocity is crucial in neuroscience and the diagnosis of cerebrovascular diseases. Traditional color Doppler ultrasound can only measure the velocity component along the ultrasound beam, which restricts its ability to accurately capture the complete blood flow vector in complex environments. To overcome these limitations, we propose an ultrafast pulse-coded vector Doppler (PC-UVD) imaging method, utilizing Hadamard matrix-based pulse encoding to improve velocity estimation accuracy under low signal-to-noise ratio (SNR) conditions. Our study encompasses spiral flow simulations and in vivo rat brain experiments, showing significantly enhanced measurement precision compared to conventional ultrafast vector Doppler (UVD). This innovative approach enables the measurement of dynamic cerebral blood flow velocity within a single cardiac cycle, offering insights into the characteristics of cerebrovascular resistivity. The proposed PC-UVD method employs Hadamard matrix encoding of plane waves, boosting SNR without compromising temporal or spatial resolution. Velocity vectors are subsequently estimated using a weighted least squares (WLS) approach, with iterative residual-based weight optimization improving robustness to noise and minimizing the impact of outliers. The effectiveness of this technique is confirmed through simulations with a spiral blood flow phantom, demonstrating a marked improvement in velocity estimation accuracy, particularly in deep imaging regions with significant signal attenuation. In vivo experiments on rat brains further confirm that the proposed method offers greater accuracy than existing UVD approaches, particularly for small vessels. Notably, our method can precisely differentiate arterial from venous flow by analyzing pulsatility and resistivity within the cerebral vascular network.

physics.med-ph↗

arXiv2024

HPL-ESS: Hybrid Pseudo-Labeling for Unsupervised Event-based Semantic Segmentation

Event-based semantic segmentation has gained popularity due to its capability to deal with scenarios under high-speed motion and extreme lighting conditions, which cannot be addressed by conventional RGB cameras. Since it is hard to annotate event data, previous approaches rely on event-to-image reconstruction to obtain pseudo labels for training. However, this will inevitably introduce noise, and learning from noisy pseudo labels, especially when generated from a single source, may reinforce the errors. This drawback is also called confirmation bias in pseudo-labeling. In this paper, we propose a novel hybrid pseudo-labeling framework for unsupervised event-based semantic segmentation, HPL-ESS, to alleviate the influence of noisy pseudo labels. In particular, we first employ a plain unsupervised domain adaptation framework as our baseline, which can generate a set of pseudo labels through self-training. Then, we incorporate offline event-to-image reconstruction into the framework, and obtain another set of pseudo labels by predicting segmentation maps on the reconstructed images. A noisy label learning strategy is designed to mix the two sets of pseudo labels and enhance the quality. Moreover, we propose a soft prototypical alignment module to further improve the consistency of target domain features. Extensive experiments show that our proposed method outperforms existing state-of-the-art methods by a large margin on the DSEC-Semantic dataset (+5.88% accuracy, +10.32% mIoU), which even surpasses several supervised methods.

cs.CV↗

arXiv2024

FairSync: Ensuring Amortized Group Exposure in Distributed Recommendation Retrieval

In pursuit of fairness and balanced development, recommender systems (RS) often prioritize group fairness, ensuring that specific groups maintain a minimum level of exposure over a given period. For example, RS platforms aim to ensure adequate exposure for new providers or specific categories of items according to their needs. Modern industry RS usually adopts a two-stage pipeline: stage-1 (retrieval stage) retrieves hundreds of candidates from millions of items distributed across various servers, and stage-2 (ranking stage) focuses on presenting a small-size but accurate selection from items chosen in stage-1. Existing efforts for ensuring amortized group exposures focus on stage-2, however, stage-1 is also critical for the task. Without a high-quality set of candidates, the stage-2 ranker cannot ensure the required exposure of groups. Previous fairness-aware works designed for stage-2 typically require accessing and traversing all items. In stage-1, however, millions of items are distributively stored in servers, making it infeasible to traverse all of them. How to ensure group exposures in the distributed retrieval process is a challenging question. To address this issue, we introduce a model named FairSync, which transforms the problem into a constrained distributed optimization problem. Specifically, FairSync resolves the issue by moving it to the dual space, where a central node aggregates historical fairness data into a vector and distributes it to all servers. To trade off the efficiency and accuracy, the gradient descent technique is used to periodically update the parameter of the dual vector. The experiment results on two public recommender retrieval datasets showcased that FairSync outperformed all the baselines, achieving the desired minimum level of exposures while maintaining a high level of retrieval accuracy.

cs.IR↗

arXiv2023

Topological expansive Lorenz maps with a hole at critical point

Let $f$ be an expansive Lorenz map and $c$ be the critical point. The survivor set is denoted as $S_{f}(H):=\{x\in[0,1]: f^{n}(x)\notin H, \forall n\geq 0\}$, where $H$ is a open subinterval. Here we study the hole $H=(a,b)$ with $a\leq c \leq b$ and $a\neq b $. We show that the case $a=c$ is equivalent to the hole at $0$, the case $b=c$ equals to the hole at $1$. We also obtain that, given an expansive Lorenz map $f$ with a hole $H=(a,b)$ and $S_{f}(H)\nsubseteqq\{0,1\}$, then there exists a Lorenz map $g$ such that $\tilde{S}_{f}(H)\setminus\Omega(g)$ is countable, where $\Omega(g)$ is the Lorenz-shift of $g$ and $\tilde{S}_{f}(H)$ is the symbolic representation of $S_{f}(H)$. Let $a$ be fixed and $b$ varies in $(c,1)$, we also give a complete characterization of the maximal interval $I(b)$ such that for all $\epsilon\in I(b)$, $S_{f}(a,\epsilon)=S_{f}(a,b)$, and $I(b)$ may degenerate to a single point $b$. Moreover, when $f$ has an ergodic acim, we show that the topological entropy function $\lambda_{f}(a):b\mapsto h_{top}(f|S_{f}(a,b))$ is a devil staircase with $a$ being fixed, so is $\lambda_{f}(b)$ if we fix $b$. At the special case $f$ being intermediate $\beta$-transformation, using the Ledrappier-Young formula, we obtain that the Hausdorff dimension function $\eta_{f}(a):b\mapsto \dim_{\mathcal{H}}(S_{f}(a,b))$ is a devil staircase when fixing $a$, so is $\eta_{f}(b)$ if $b$ is fixed. As a result, we extend the devil staircases in \cite{Urbanski1986,kalle2020,Langeveld2023} to expansive Lorenz maps with a hole at critical point.

math.DS↗

arXiv2022

Subshifts of finite type and matching for intermediate $\beta$-transformations

We focus on the relationships between matching and subshift of finite type for intermediate $\beta$-transformations $T_{\beta,\alpha}(x)=\beta x+\alpha $ ($\bmod$ 1), where $x\in[0,1]$ and $(\beta,\alpha) \in \Delta:= \{ (\beta, \alpha) \in \mathbb{R}^{2}:\beta \in (1, 2) \; \rm{and} \; 0 < \alpha <2 - \beta\}$. We prove that if the kneading space $\Omega_{\beta,\alpha}$ is a subshift of finite type, then $T_{\beta,\alpha}$ has matching. Moreover, each $(\beta,\alpha)\in\Delta$ with $T_{\beta,\alpha}$ has matching corresponds to a matching interval, and there are at most countable different matching intervals on the fiber. Using combinatorial approach, we construct a pair of linearizable periodic kneading invariants and show that, for any $\epsilon>0$ and $(\beta,\alpha)\in\Delta$ with $T_{\beta,\alpha}$ has matching, there exists $(\beta,\alpha^{\prime})$ on the fiber with $|\alpha-\alpha^{\prime}|<\epsilon$, such that $\Omega_{\beta,\alpha^{\prime}}$ is a subshift of finite type. As a result, the set of $(\beta,\alpha)$ for which $\Omega_{\beta,\alpha}$ is a subshift of finite type is dense on the fiber if and only if the set of $(\beta,\alpha)$ for which $T_{\beta,\alpha}$ has matching is dense on the fiber.

math.DS↗

arXiv2022

Thick hyperbolic repelling invariant Cantor set and wild attractor

Let $D$ be the set of $\beta \in (1, 2]$ such that $f_\beta$ is a symmetric tent map with finite critical orbit. For $\beta \in D$, by operating Denjoy like surgery on $f_{\beta}$, we constructed a $C^1$ unimodal map $\tilde{g}_\beta$ admitting a thick hyperbolic repelling invariant Cantor set which contains a wild Cantor attractor. The smoothness of $\tilde{g}_\beta$ is ensured by the effective estimation of the preimages of the critical point as well as the prescribed lengths of the inserted intervals. Furthermore, $D$ is dense in $(1, 2]$, and $\tilde{g}_\beta$ can not be $C^{1+\alpha}$ because the hyperbolic repelling invariant Cantor set of $C^{1+\alpha}$ map has Lebesgue measure equal to zero.

math.DS↗

arXiv2021

Digital quantum simulation and Pseudoquantum Simulation of $\mathbb{Z}_2$ Gauge Higgs Model

We present a quantum algorithm for digital quantum simulation of the $\mathbb{Z}_2$ gauge-Higgs model on a $3\times 3$ lattice, which is based on Trotter decomposition, the quantum adiabatic algorithm and its circuit realization. Then we perform a classical demonstration, dubbed a pseudoquantum simulation, on a GPU simulator. We obtain useful results on this model, which suggest the topological properties of the deconfined phase and help to clarify the phase diagram. It is suggested that the tricitical point, where the second-order critical lines of deconfinement-confinement transition and of deconfinement-Higgs transition meet, seems to be on the the first-order critical line of confinement-Higgs transition, at a point other than the end of this critical line.

hep-lat↗

arXiv2021

Complete invariants and parametrization of expansive Lorenz maps

We obtain the complete conjugacy invariants of expansive Lorenz maps and for any given two expansive Lorenz maps, there are two unique sequences of $(β_{i},α_{i})$ pairs. In this way, we can define the classification of expansive Lorenz maps. Moreover, we investigate the uniform linearization of expansive Lorenz maps through periodic renormalization.

math.DS↗

arXiv2021

REFIT: A Unified Watermark Removal Framework For Deep Learning Systems With Limited Data

Training deep neural networks from scratch could be computationally expensive and requires a lot of training data. Recent work has explored different watermarking techniques to protect the pre-trained deep neural networks from potential copyright infringements. However, these techniques could be vulnerable to watermark removal attacks. In this work, we propose REFIT, a unified watermark removal framework based on fine-tuning, which does not rely on the knowledge of the watermarks, and is effective against a wide range of watermarking schemes. In particular, we conduct a comprehensive study of a realistic attack scenario where the adversary has limited training data, which has not been emphasized in prior work on attacks against watermarking schemes. To effectively remove the watermarks without compromising the model functionality under this weak threat model, we propose two techniques that are incorporated into our fine-tuning framework: (1) an adaption of the elastic weight consolidation (EWC) algorithm, which is originally proposed for mitigating the catastrophic forgetting phenomenon; and (2) unlabeled data augmentation (AU), where we leverage auxiliary unlabeled data from other sources. Our extensive evaluation shows the effectiveness of REFIT against diverse watermark embedding schemes. In particular, both EWC and AU significantly decrease the amount of labeled training data needed for effective watermark removal, and the unlabeled data samples used for AU do not necessarily need to be drawn from the same distribution as the benign data for model evaluation. The experimental results demonstrate that our fine-tuning based watermark removal attacks could pose real threats to the copyright of pre-trained models, and thus highlight the importance of further investigating the watermarking problem and proposing more robust watermark embedding schemes against the attacks.

cs.CR↗

arXiv2020

{\alpha}-limit sets and Lyapunov function for maps with one topological attractor

We consider the topological behaviors of continuous maps with one topological attractor on compact metric space $X$. This kind of map is a generalization of maps such as topologically expansive Lorenz map, unimodal map without homtervals and so on. We provide a leveled $A$-$R$ pair decomposition for such maps, and characterize $\alpha$-limit set of each point. Based on weak Morse decomposition of $X$, we construct a bounded Lyapunov function $V(x)$, which give a clear description of orbit behavior of each point in $X$ except a meager set.

math.DS↗

arXiv2020

Goal-conditioned Imitation Learning

Designing rewards for Reinforcement Learning (RL) is challenging because it needs to convey the desired task, be efficient to optimize, and be easy to compute. The latter is particularly problematic when applying RL to robotics, where detecting whether the desired configuration is reached might require considerable supervision and instrumentation. Furthermore, we are often interested in being able to reach a wide range of configurations, hence setting up a different reward every time might be unpractical. Methods like Hindsight Experience Replay (HER) have recently shown promise to learn policies able to reach many goals, without the need of a reward. Unfortunately, without tricks like resetting to points along the trajectory, HER might require many samples to discover how to reach certain areas of the state-space. In this work we investigate different approaches to incorporate demonstrations to drastically speed up the convergence to a policy able to reach any goal, also surpassing the performance of an agent trained with other Imitation Learning algorithms. Furthermore, we show our method can also be used when the available expert trajectories do not contain the actions, which can leverage kinesthetic or third person demonstration. The code is available at https://sites.google.com/view/goalconditioned-il/.

cs.LG↗

arXiv2020

Mutual Information Maximization for Robust Plannable Representations

Extending the capabilities of robotics to real-world complex, unstructured environments requires the need of developing better perception systems while maintaining low sample complexity. When dealing with high-dimensional state spaces, current methods are either model-free or model-based based on reconstruction objectives. The sample inefficiency of the former constitutes a major barrier for applying them to the real-world. The later, while they present low sample complexity, they learn latent spaces that need to reconstruct every single detail of the scene. In real environments, the task typically just represents a small fraction of the scene. Reconstruction objectives suffer in such scenarios as they capture all the unnecessary components. In this work, we present MIRO, an information theoretic representational learning algorithm for model-based reinforcement learning. We design a latent space that maximizes the mutual information with the future information while being able to capture all the information needed for planning. We show that our approach is more robust than reconstruction objectives in the presence of distractors and cluttered scenes

cs.CV↗

arXiv2020

Dual-comb spectroscopy for high-temperature reaction kinetics

In the current study, a quantum-cascade-laser-based dual-comb spectrometer (DCS) was used to paint a detailed picture of a 1.0 ms high-temperature reaction between propyne and oxygen. The DCS interfaced with a shock tube to provide pre-ignition conditions of 1225 K, 2.8 atm, and 2% p-C3H4/18% O2/Ar. The spectrometer consisted of two free-running, non-stabilized frequency combs each emitting at 179 wavelengths between 1174 and 1233 cm-1. A free spectral range, f_r, of 9.86 GHz and a difference in comb spacing, Δf_r, of 5 MHz, enabled a theoretical time resolution of 0.2 us but the data was time-integrated to 4 us to improve SNR. The accuracy of the spectrometer was monitored using a suite of independent laser diagnostics and good agreement observed.

physics.chem-ph↗

arXiv2019

Information-Driven Design for Shock Tube / Laser Absorption Studies of Fundamental Rate Constants in Combustion, with Application to Methanol Pyrolysis

Shock tube experiments, paired with precision laser diagnostics, are ideal venues to provide kinetics data critically needed for the development, validation and optimization of modern combustion kinetics models. However, to design sensitive, accurate, feasible and information-rich experiments that may yield such data often requires sophisticated planning. This study presents a mathematical framework and quantitative approach to guide such experimental design, namely a method to pin-point the optimal conditions for specific experimentation under realistic constraints of the shock tubes and diagnostic tools involved. For demonstration purpose, the current work is focused on a key type of shock tube kinetic experiments -- direct determination of fundamental reaction rate constants. Specifically, this study utilizes a Bayesian approach to maximize the prior-posterior gain in Shannon information of the rate constants to be inferred from the intended experiment. Example application of this method to the experimental determination of the CH$_3$OH + H (k$_1$) and CH$_2$O + H (k$_2$) rate constants is demonstrated in shock tube/laser absorption studies of the CH$_3$OH pyrolysis system, yielding new recommended rate constant expressions (over 1287 K - 1537 K) as: k$_1$ = $ 2.50 \times 10^6 (T/K)^{2.35} exp(-2975 K /T) \, cm^3mol^{-1}s^{-1} \pm 11.4\%$ and k$_2$ = $7.06 \times 10^7 (T/K)^{1.9} exp(-1380 K/T) \, cm^3mol^{-1}s^{-1} \pm 9.7 \%$. Potential extension to other types of kinetic studies, e.g. prediction of combustion benchmarks such as ignition delay times and species yields, and global uncertainty minimization of generic reaction models, are also briefly discussed.

physics.chem-ph↗

arXiv2018

Equitability of Dependence Measure

Measuring dependence between two random variables is very important, and critical in many applied areas such as variable selection, brain network analysis. However, we do not know what kind of functional relationship is between two covariates, which requires the dependence measure to be equitable. That is, it gives similar scores to equally noisy relationship of different types. In fact, the dependence score is a continuous random variable taking values in $[0,1]$, thus it is theoretically impossible to give similar scores. In this paper, we introduce a new definition of equitability of a dependence measure, i.e, power-equitable (weak-equitable) and show by simulation that HHG and Copula Dependence Coefficient (CDC) are weak-equitable.

stat.ML↗