This paper considers the open challenge of identifying complete, concise, and explainable quantitative microstructure representations for disordered heterogeneous material systems. Completeness and conciseness have been achieved through existing data
-driven methods, e.g., deep generative models, which, however, do not provide mathematically explainable latent representations. This study investigates representations composed of three-point correlation functions, which are a special type of spatial convolutions. We show that a variety of microstructures can be characterized by a concise subset of three-point correlations, and the identification of such subsets can be achieved by Bayesian optimization. Lastly, we show that the proposed representation can directly be used to compute material properties based on the effective medium theory.
Optical flow estimation with occlusion or large displacement is a problematic challenge due to the lost of corresponding pixels between consecutive frames. In this paper, we discover that the lost information is related to a large quantity of motion
features (more than 40%) computed from the popular discriminative cost-volume feature would completely vanish due to invalid sampling, leading to the low efficiency of optical flow learning. We call this phenomenon the Vanishing Cost Volume Problem. Inspired by the fact that local motion tends to be highly consistent within a short temporal window, we propose a novel iterative Motion Feature Recovery (MFR) method to address the vanishing cost volume via modeling motion consistency across multiple frames. In each MFR iteration, invalid entries from original motion features are first determined based on the current flow. Then, an efficient network is designed to adaptively learn the motion correlation to recover invalid features for lost-information restoration. The final optical flow is then decoded from the recovered motion features. Experimental results on Sintel and KITTI show that our method achieves state-of-the-art performances. In fact, MFR currently ranks second on Sintel public website.
In 2D+3D facial expression recognition (FER), existing methods generate multi-view geometry maps to enhance the depth feature representation. However, this may introduce false estimations due to local plane fitting from incomplete point clouds. In th
is paper, we propose a novel Map Generation technique from the viewpoint of information theory, to boost the slight 3D expression differences from strong personality variations. First, we examine the HDR depth data to extract the discriminative dynamic range $r_{dis}$, and maximize the entropy of $r_{dis}$ to a global optimum. Then, to prevent the large deformation caused by over-enhancement, we introduce a depth distortion constraint and reduce the complexity from $O(KN^2)$ to $O(KNtau)$. Furthermore, the constrained optimization is modeled as a $K$-edges maximum weight path problem in a directed acyclic graph, and we solve it efficiently via dynamic programming. Finally, we also design an efficient Facial Attention structure to automatically locate subtle discriminative facial parts for multi-scale learning, and train it with a proposed loss function $mathcal{L}_{FA}$ without any facial landmarks. Experimental results on different datasets show that the proposed method is effective and outperforms the state-of-the-art 2D+3D FER methods in both FER accuracy and the output entropy of the generated maps.
This paper addresses the challenging unsupervised scene flow estimation problem by jointly learning four low-level vision sub-tasks: optical flow $textbf{F}$, stereo-depth $textbf{D}$, camera pose $textbf{P}$ and motion segmentation $textbf{S}$. Our
key insight is that the rigidity of the scene shares the same inherent geometrical structure with object movements and scene depth. Hence, rigidity from $textbf{S}$ can be inferred by jointly coupling $textbf{F}$, $textbf{D}$ and $textbf{P}$ to achieve more robust estimation. To this end, we propose a novel scene flow framework named EffiScene with efficient joint rigidity learning, going beyond the existing pipeline with independent auxiliary structures. In EffiScene, we first estimate optical flow and depth at the coarse level and then compute camera pose by Perspective-$n$-Points method. To jointly learn local rigidity, we design a novel Rigidity From Motion (RfM) layer with three principal components: emph{}{(i)} correlation extraction; emph{}{(ii)} boundary learning; and emph{}{(iii)} outlier exclusion. Final outputs are fused based on the rigid map $M_R$ from RfM at finer levels. To efficiently train EffiScene, two new losses $mathcal{L}_{bnd}$ and $mathcal{L}_{unc}$ are designed to prevent trivial solutions and to regularize the flow boundary discontinuity. Extensive experiments on scene flow benchmark KITTI show that our method is effective and significantly improves the state-of-the-art approaches for all sub-tasks, i.e. optical flow ($5.19 rightarrow 4.20$), depth estimation ($3.78 rightarrow 3.46$), visual odometry ($0.012 rightarrow 0.011$) and motion segmentation ($0.57 rightarrow 0.62$).
Cell migration, which can be significantly affected by intracellular signaling pathways (ICSP) and extracellular matrix (ECM), plays a crucial role in many physiological and pathological processes. The efficiency of cell migration, which is typically
modeled as a persistent random walk (PRW), depends on two critical motility parameters, i.e., migration speed and persistence. It is generally very challenging to efficiently and accurately extract these key dynamics parameters from noisy experimental data. Here, we employ the normalized Shannon entropy to quantify the deviation of cell migration dynamics from that of diffusive/ballistic motion as well as to derive the persistence of cell migration based on the Fourier power spectrum of migration velocities. Moreover, we introduce the time-varying Shannon entropy based on the wavelet power spectrum of cellular dynamics and demonstrate its superior utility to characterize the time-dependent persistence of cell migration, which is typically resulted from complex and time-varying intra or extra-cellular mechanisms. We employ our approach to analyze trajectory data of in vitro cell migration regulated by distinct intracellular and extracellular mechanisms, exhibiting a rich spectrum of dynamic characteristics. Our analysis indicates that the combination of Shannon entropy and wavelet transform offers a simple and efficient tool to estimate the persistence of cell migration, which may also reflect the real-time effects of ICSP-ECM to some extent.
Cell migration is an indispensable physiological and pathological process for normal tissue development and cancer metastasis, which is greatly regulated by intracellular signal pathways and extracellular microenvironment (ECM). However, there is a l
ack of adequate tools to analyze the time-varying cell migration characteristics because of the effects of some factors, i.e., the ECM including the time-dependent local stiffness due to microstructural remodeling by migrating cells. Here, we develop an approach to derive the time-dependent motility parameters from cellular trajectories, based on the time-varying persistent random walk model. In particular, we employ the wavelet denoising and wavelet transform to investigate cell migration velocities and obtain the wavelet power spectrum. The time-dependent motility parameters are subsequently derived via Lorentzian power spectrum. Our analysis shows that the combination of wavelet denoising, wavelet transform and Lorentzian power spectrum provides a powerful tool to derive accurately the time-dependent motility parameters, which reflects the time-varying microenvironment characteristics to some extent.
Multivariate time series (MTS) data are becoming increasingly ubiquitous in diverse domains, e.g., IoT systems, health informatics, and 5G networks. To obtain an effective representation of MTS data, it is not only essential to consider unpredictable
dynamics and highly variable lengths of these data but also important to address the irregularities in the sampling rates of MTS. Existing parametric approaches rely on manual hyperparameter tuning and may cost a huge amount of labor effort. Therefore, it is desirable to learn the representation automatically and efficiently. To this end, we propose an autonomous representation learning approach for multivariate time series (TimeAutoML) with irregular sampling rates and variable lengths. As opposed to previous works, we first present a representation learning pipeline in which the configuration and hyperparameter optimization are fully automatic and can be tailored for various tasks, e.g., anomaly detection, clustering, etc. Next, a negative sample generation approach and an auxiliary classification task are developed and integrated within TimeAutoML to enhance its representation capability. Extensive empirical studies on real-world datasets demonstrate that the proposed TimeAutoML outperforms competing approaches on various tasks by a large margin. In fact, it achieves the best anomaly detection performance among all comparison algorithms on 78 out of all 85 UCR datasets, acquiring up to 20% performance improvement in terms of AUC score.
We propose a scheme to realize optical nonreciprocal response and conversion in a Tavis-Cummings coupling optomechanical system, where a single cavity mode interacts with the vibrational mode of a flexible membrane with an embedded ensemble of two-le
vel quantum emitters. Due to the introduction of the Tavis-Cummings interaction, we find that the phases between the mechanical mode and the optical mode, as well as between the mechanical mode and the dopant mode, are correlated with each other, and further give the analytical relationship between them. By optimizing the system parameters, especially the relative phase between two paths, the optimal nonreciprocal response can be achieved. Under the frequency domain, we derive the transmission matrix of the system analytically based on the input-output relation and study the influence of the system parameters on the nonreciprocal response of the quantum input signal. Moreover, compared with the conventional optomechanical systems, the Tavis-Cummings coupling optomechanical system exhibits richer nonreciprocal conversion phenomena among the optical mode, mechanical mode, and dopant mode, which provide a new applicable way of achieving the phonon-photon transducer and the optomechanical circulator in future practice.
The van der Waals (vdW) density functional (vdW-DF) method [ROPP 78, 066501 (2015)] describes dispersion or vdW binding by tracking the effects of an electrodynamic coupling among pairs of electrons and their associated exchange-correlation holes. Th
is is done in a nonlocal-correlation energy term $E_c^{nl}$, which permits density functional theory calculation in the Kohn-Sham scheme. However, to map the nature of vdW forces in the fully interacting materials system, it is necessary to compensate for associated kinetic-correlation energy effects. Here we present a coupling-constant scaling analysis that also permits us to compute the kinetic-correlation energy $T_c^{nl}$ that is specific to the vdW-DF account of nonlocal correlations. We thus provide a spatially-resolved analysis of the total nonlocal-correlation binding, including vdW forces, in both covalently and non-covalently bonded systems. We find that kinetic-correlation energy effects play a significant role in the account of vdW or dispersion interactions among molecules. We also find that the signatures that we reveal in our full-interaction mapping are typically given by the spatial variation in the $E_c^{nl}$ binding contributions, at least in a qualitative discussion. Furthermore, our full mapping shows that the total nonlocal-correlation binding is concentrated to pockets in the sparse electron distribution located between the material fragments.
Two hybrid van der Waals density functionals (vdW-DFs) are constructed using 25%, Fock exchange with i) the consistent-exchange vdW-DF-cx functional and ii) with the vdW-DF2 functional. The ability to describe covalent and non-covalent binding proper
ties of molecules are assessed. For properties related to covalent binding, atomization energies (G2-1 set), molecular reaction energies (G2RC set), as well as ionization energies (G21IP set) are benchmarked against experimental reference values. We find that hybrid-vdW-DF-cx yields results that are rather similar to those of the standard non-empirical hybrid PBE0 [JCP 110, 6158 (1996)]. Hybrid vdW-DF2 follows somewhat different trends, showing on average significantly larger deviations from the reference energies, with a MAD of 14.5 kcal/mol for the G2-1 set. Non-covalent binding properties of molecules are assessed using the S22 benchmark set of non-covalently bonded dimers and the X40 set of dimers of small halogenated molecules, using wavefunction-based quantum chemistry results for references. For the S22 set, hybrid-vdW-DF-cx performs better than standard vdW-DF-cx for the mostly hydrogen-bonded systems. Hybrid-vdW-DF2 offers a slight improvement over standard vdW-DF2. Similar trends are found for the X40 set, with hybrid-vdW-DF-cx performing particularly well for binding involving the strongly polar hydrogen halides, but poorly for systems with tiny binding energies. Our study of the X40 set reveals both the potential of mixing Fock exchange with vdW-DF, but also highlights shortcomings of the hybrids constructed here. The solid performance of hybrid-vdW-DF-cx for covalent-bonded systems, as well as the strengths and issues uncovered for non-covalently bonded systems, makes this study a good starting point for developing even more precise hybrid vdW-DFs.
