arXiv subjects

Ding Ding

Publications and source records attributed to Ding Ding.

At least 37 records · Page 2Linked to original sources

arXiv2024

Study on the Influence of Embodied Avatars on Gait Parameters in Virtual Environments and Real World

In this study, we compare the virtual and real gait parameters to investigate the effect of appearances of embodied avatars and virtual reality experience on gait in physical and virtual environments. We developed a virtual environment simulation and gait detection system for analyzing gait. The system transfers real-life scenarios into a realistic presentation in the virtual environment and provides look-alike same-age and old-age avatars for participants. We conducted an empirical study and used subjective questionnaires to evaluate participants' feelings about the virtual reality experience. Also, the paired sample t-test and neural network were implemented to analyze gait differences. The results suggest that there are disparities in gait between virtual and real environments. Also, the appearance of embodied avatars could influence the gait parameters in the virtual environment. Moreover, the experience of embodying old-age avatars affects the gait in the real world.

cs.HC↗

arXiv2024

Transferable Learned Image Compression-Resistant Adversarial Perturbations

Adversarial attacks can readily disrupt the image classification system, revealing the vulnerability of DNN-based recognition tasks. While existing adversarial perturbations are primarily applied to uncompressed images or compressed images by the traditional image compression method, i.e., JPEG, limited studies have investigated the robustness of models for image classification in the context of DNN-based image compression. With the rapid evolution of advanced image compression, DNN-based learned image compression has emerged as the promising approach for transmitting images in many security-critical applications, such as cloud-based face recognition and autonomous driving, due to its superior performance over traditional compression. Therefore, there is a pressing need to fully investigate the robustness of a classification system post-processed by learned image compression. To bridge this research gap, we explore the adversarial attack on a new pipeline that targets image classification models that utilize learned image compressors as pre-processing modules. Furthermore, to enhance the transferability of perturbations across various quality levels and architectures of learned image compression models, we introduce a saliency score-based sampling method to enable the fast generation of transferable perturbation. Extensive experiments with popular attack methods demonstrate the enhanced transferability of our proposed method when attacking images that have been post-processed with different learned image compression models.

cs.CV↗

arXiv2024

Meso-inflationary Peccei-Quinn symmetry breaking with non-minimal coupling

We study a realization of the inflationary scenario where the Peccei-Quinn (PQ) symmetry is spontaneously broken during inflation, facilitated by its non-minimal coupling to gravity. This results in effectively two-field inflation: the early stage is driven by an inflaton field with the PQ symmetry intact, and the later stage is driven by the PQ scalar after its effective mass becomes tachyonic, causing destabilization from the origin. The non-minimal coupling serves the dual purpose of restoring the PQ symmetry during early inflation and flattening the PQ potential post-tachyonic shift, allowing for continued slow roll. We analyze the inflationary background solutions and scalar perturbations, which are amplified at small scales via significant isocurvature perturbations generated near the symmetry-breaking epoch. These perturbations lead to second-order gravitational waves, detectable by next-generation space-based experiments.

hep-th↗

arXiv2024

Effective dynamics of quantum fluctuations in field theory: with applications to cosmology

We develop a novel framework for describing quantum fluctuations in field theory, with a focus on cosmological applications. Our method uniquely circumvents the use of operator/Hilbert-space formalism, instead relying on a systematic treatment of classical variables, quantum fluctuations, and an effective Hamiltonian. Our framework not only aligns with standard formalisms in flat and de Sitter spacetimes, which assumes no backreaction, demonstrated through the $φ^3$-model, but also adeptly handles time-dependent backreaction in more general cases. The uncertainty principle and spatial symmetry emerge as critical tools for selecting initial conditions and understanding effective potentials. We discover that modes inside the Hubble horizon \emph{do not} necessarily feel an initial Minkowski vacuum, as is commonly assumed. Our findings offer fresh insights into the early universe's quantum fluctuations and potential explanations to large-scale CMB anomalies.

gr-qc↗

arXiv2023

Corner-to-Center Long-range Context Model for Efficient Learned Image Compression

In the framework of learned image compression, the context model plays a pivotal role in capturing the dependencies among latent representations. To reduce the decoding time resulting from the serial autoregressive context model, the parallel context model has been proposed as an alternative that necessitates only two passes during the decoding phase, thus facilitating efficient image compression in real-world scenarios. However, performance degradation occurs due to its incomplete casual context. To tackle this issue, we conduct an in-depth analysis of the performance degradation observed in existing parallel context models, focusing on two aspects: the Quantity and Quality of information utilized for context prediction and decoding. Based on such analysis, we propose the \textbf{Corner-to-Center transformer-based Context Model (C$^3$M)} designed to enhance context and latent predictions and improve rate-distortion performance. Specifically, we leverage the logarithmic-based prediction order to predict more context features from corner to center progressively. In addition, to enlarge the receptive field in the analysis and synthesis transformation, we use the Long-range Crossing Attention Module (LCAM) in the encoder/decoder to capture the long-range semantic information by assigning the different window shapes in different channels. Extensive experimental evaluations show that the proposed method is effective and outperforms the state-of-the-art parallel methods. Finally, according to the subjective analysis, we suggest that improving the detailed representation in transformer-based image compression is a promising direction to be explored.

eess.IV↗

arXiv2023

Reconstruction Distortion of Learned Image Compression with Imperceptible Perturbations

Learned Image Compression (LIC) has recently become the trending technique for image transmission due to its notable performance. Despite its popularity, the robustness of LIC with respect to the quality of image reconstruction remains under-explored. In this paper, we introduce an imperceptible attack approach designed to effectively degrade the reconstruction quality of LIC, resulting in the reconstructed image being severely disrupted by noise where any object in the reconstructed images is virtually impossible. More specifically, we generate adversarial examples by introducing a Frobenius norm-based loss function to maximize the discrepancy between original images and reconstructed adversarial examples. Further, leveraging the insensitivity of high-frequency components to human vision, we introduce Imperceptibility Constraint (IC) to ensure that the perturbations remain inconspicuous. Experiments conducted on the Kodak dataset using various LIC models demonstrate effectiveness. In addition, we provide several findings and suggestions for designing future defenses.

cs.CV↗

arXiv2022

BigDL 2.0: Seamless Scaling of AI Pipelines from Laptops to Distributed Cluster

Most AI projects start with a Python notebook running on a single laptop; however, one usually needs to go through a mountain of pains to scale it to handle larger dataset (for both experimentation and production deployment). These usually entail many manual and error-prone steps for the data scientists to fully take advantage of the available hardware resources (e.g., SIMD instructions, multi-processing, quantization, memory allocation optimization, data partitioning, distributed computing, etc.). To address this challenge, we have open sourced BigDL 2.0 at https://github.com/intel-analytics/BigDL/ under Apache 2.0 license (combining the original BigDL and Analytics Zoo projects); using BigDL 2.0, users can simply build conventional Python notebooks on their laptops (with possible AutoML support), which can then be transparently accelerated on a single node (with up-to 9.6x speedup in our experiments), and seamlessly scaled out to a large cluster (across several hundreds servers in real-world use cases). BigDL 2.0 has already been adopted by many real-world users (such as Mastercard, Burger King, Inspur, etc.) in production.

cs.LG↗

arXiv2021

Canonical description of cosmological backreaction

Canonical methods of quasiclassical dynamics make it possible to go beyond a strict background approximation for cosmological perturbations by including independent fields such as correlation degrees of freedom. New models are introduced and analyzed here for cosmological dynamics in the presence of quantum correlations between background and perturbations, as well as cross-correlations between different modes of a quantum field. Evolution equations for moments of a perturbation state reveal conditions required for inhomogeneity to build up out of an initial vacuum. A crucial role is played by quantum non-locality, formulated by canonical methods as an equivalent local theory with non-classical degrees of freedom given by moments of a quantum state.

gr-qc↗

arXiv2021

Polyjuice: High-Performance Transactions via Learned Concurrency Control

Concurrency control algorithms are key determinants of the performance of in-memory databases. Existing algorithms are designed to work well for certain workloads. For example, optimistic concurrency control (OCC) is better than two-phase-locking (2PL) under low contention, while the converse is true under high contention. To adapt to different workloads, prior works mix or switch between a few known algorithms using manual insights or simple heuristics. We propose a learning-based framework that instead explicitly optimizes concurrency control via offline training to maximize performance. Instead of choosing among a small number of known algorithms, our approach searches in a "policy space" of fine-grained actions, resulting in novel algorithms that can outperform existing algorithms by specializing to a given workload. We build Polyjuice based on our learning framework and evaluate it against several existing algorithms. Under different configurations of TPC-C and TPC-E, Polyjuice can achieve throughput numbers higher than the best of existing algorithms by 15% to 56%.

cs.DB↗

arXiv2020

A Graph Attention Based Approach for Trajectory Prediction in Multi-agent Sports Games

This work investigates the problem of multi-agents trajectory prediction. Prior approaches lack of capability of capturing fine-grained dependencies among coordinated agents. In this paper, we propose a spatial-temporal trajectory prediction approach that is able to learn the strategy of a team with multiple coordinated agents. In particular, we use graph-based attention model to learn the dependency of the agents. In addition, instead of utilizing the recurrent networks (e.g., VRNN, LSTM), our method uses a Temporal Convolutional Network (TCN) as the sequential model to support long effective history and provide important features such as parallelism and stable gradients. We demonstrate the validation and effectiveness of our approach on two different sports game datasets: basketball and soccer datasets. The result shows that compared to related approaches, our model that infers the dependency of players yields substantially improved performance. Code is available at https://github.com/iHeartGraph/predict

cs.LG↗

arXiv2020

Multi-field inflation from single-field models

Quantization implies independent degrees of freedom that do not appear in the classical theory, given by fluctuations, correlations, and higher moments of a state. A systematic derivation of the resulting dynamical systems is presented here in a cosmological application for near-Gaussian states of a single-field inflation model. As a consequence, single-field Higgs inflation is made viable observationally by becoming a multi-field model with a specific potential for a fluctuation field interacting with the inflaton expectation value. Crucially, non-adiabatic methods of semiclassical quantum dynamics reveal important phases that can set suitable initial conditions for slow-roll inflation (in combination with the uncertainty relation), and then end inflation after the observationally preferred number of e-folds. New parameters in the interaction potential are derived from properties of the underlying background state, demonstrating how background non-Gaussianity can affect observational features of inflation or, conversely, how observations may be used to understand the quantum state of the inflaton.

hep-th↗

arXiv2020

Quantum Higgs Inflation

The Higgs field is an attractive candidate for the inflaton because it is an observationally confirmed fundamental scalar field. Importantly, it can be modeled by the most general renormalizable scalar potential. However, if the classical Higgs potential is used in models of inflation, it is ruled out by detailed observations of the cosmic microwave background. Here, a new application of non-adiabatic quantum dynamics to cosmological models is shown to lead to a multi-field Higgs-like potential, consistent with observations of a slightly red-tilted power spectrum and a small tensor-to-scalar ratio, without requiring non-standard ingredients. These methods naturally lead to novel effects in the beginning of inflation, circumventing common fine-tuning issues by an application of uncertainty relations to estimate the initial quantum fluctuations in the early universe. Moreover, inflation ends smoothly as a consequence of the derived multi-field interactions.

astro-ph.CO↗

arXiv2019

BigDL: A Distributed Deep Learning Framework for Big Data

This paper presents BigDL (a distributed deep learning framework for Apache Spark), which has been used by a variety of users in the industry for building deep learning applications on production big data platforms. It allows deep learning applications to run on the Apache Hadoop/Spark cluster so as to directly process the production data, and as a part of the end-to-end data analysis pipeline for deployment and management. Unlike existing deep learning frameworks, BigDL implements distributed, data parallel training directly on top of the functional compute model (with copy-on-write and coarse-grained operations) of Spark. We also share real-world experience and "war stories" of users that have adopted BigDL to address their challenges(i.e., how to easily build end-to-end data analysis and deep learning pipelines for their production data).

cs.DC↗

arXiv2019

Deformed covariance in spherically symmetric vacuum models of loop quantum gravity: Consistency in Euclidean and self-dual gravity

Different versions of consistent canonical realizations of hypersurface deformations of spherically symmetric space-times have been derived in models of loop quantum gravity, modifying the classical dynamics and sometimes also the structure of space-time. Based on a canonical version of effective field theory, this paper provides a unified treatment, showing that modified space-time structures are generic in this setting. The special case of Euclidean gravity demonstrates agreement also with existing operator calculations.

gr-qc↗

arXiv2019

Hassle-free Approach to Thermal Transport Measurements Using Spatial-Temporal Temperature Data

Nanoscale engineering and novel materials have created interesting effects in thermal transport. Thermal conductivity can now be different due to physical and heating sizes. Also, highly anisotropic thermal conductivity can result from unique material composition and geometries. Various experimental methods have been developed to measure these thermal conductivity variations. All of them require varying the physical size of the sample, the heating size or relative positions between heating and detection. Here, we numerically propose a time-domain optical method that uses spatial temporal temperature data to resolve anisotropic and size-dependent thermal conductivity. Our method is hassle-free as it does not vary any experimental parameters and is easily compatible with various methods of measuring temperature in the time domain. This technique can high throughput screening of thermal properties for nanoengineered and novel materials in thermal transport. Also, this technique can be used to identify novel effects in thermal transport within a single experiment.

cond-mat.mtrl-sci↗

arXiv2019

Machine Learning for Novel Thermal-Materials Discovery: Early Successes, Opportunities, and Challenges

High-throughput computational and experimental design of materials aided by machine learning have become an increasingly important field in material science. This area of research has emerged in leaps and bounds in the thermal sciences, in part due to the advances in computational and experimental methods in obtaining thermal properties of materials. In this paper, we provide a current overview of some of the recent work and highlight the challenges and opportunities that are ahead of us in this field. In particular, we focus on the use of machine learning and high-throughput methods for screening of thermal conductivity for compounds, composites and alloys as well as interfacial thermal conductance. These new tools have brought about a feedback mechanism for understanding new correlations and identifying new descriptors, speeding up the discovery of novel thermal functional materials.

cond-mat.mtrl-sci↗

arXiv2018

Phonon Coherence and Its Effect on Thermal Conductivity of Nanostructures

The concept of coherence is one of the fundamental phenomena in electronics and optics. In addition to electron and photon, phonon, is another important energy and information carrier in nature. Without any doubt, exploration of the phonon coherence and its impact on thermal conduction will markedly change many aspects in broad applications for heat control and management in the real world. So far, the application of coherent effect on manipulation of phonon transport is a challenge work. In this article, we review recent advances in the study of the phonon coherent transport in nanomaterials and nanostructures. We first briefly look back the classical and quantum theory of coherence. Next, we review the progresses made in the understanding of phonon coherence in superlattice, nanowires and nanomeshes, respectively, focus on the effect of phonon coherence on thermal conductivity. Finally, we introduce the recent advances in the direct detection of phonon coherence using optical coherence theory.

cond-mat.mes-hall↗

arXiv2017

Lasing with cell-endogenous fluorophores: parameters and conditions

The notion of lasing with biologics has recently been realized and has since rapidly developed with the collective objective of creating lasers $\textit{in vivo}$. One limitation of achieving this goal is the requirement of exogenous laser dyes and fluorescent materials. To circumvent this, we investigate the use of cell-endogenous fluorophores - sources of cell autofluorescence - as laser gain material. In this work, we study the lasing potential and efficiency of flavins and reduced nicotinamide adenine dinucleotide (phosphate) (NAD(P)H) using a dye lasing model based on coupled rate equations. Analytical solutions for one- and two-photon pumped system were used in multi-parameter studies. We found that at physiological conditions, lasing can be supported by NAD(P)H with cavity quality factors of $10^5$. With the further consideration of damage thresholds, we recommend the use of flavins as they entail lower threshold requirements. We then identify potential parameters for engineering to make the lasing of flavins feasible even at their low physiological intracellular concentrations. We also note the higher threshold requirements and lower efficiencies of two-photon pumping, but recognize its potential for realizing lasing $\textit{in vivo}$.

physics.bio-ph↗