Searcharxiv⌕ Search

arXiv subjects

Jie Liu

Publications and source records attributed to Jie Liu.

At least 415 records · Page 23Linked to original sources
arXiv2023

CLIP-Driven Universal Model for Organ Segmentation and Tumor Detection

An increasing number of public datasets have shown a marked impact on automated organ segmentation and tumor detection. However, due to the small size and partially labeled problem of each dataset, as well as a limited investigation of diverse types of tumors, the resulting models are often limited to segmenting specific organs/tumors and ignore the semantics of anatomical structures, nor can they be extended to novel domains. To address these issues, we propose the CLIP-Driven Universal Model, which incorporates text embedding learned from Contrastive Language-Image Pre-training (CLIP) to segmentation models. This CLIP-based label encoding captures anatomical relationships, enabling the model to learn a structured feature embedding and segment 25 organs and 6 types of tumors. The proposed model is developed from an assembly of 14 datasets, using a total of 3,410 CT scans for training and then evaluated on 6,162 external CT scans from 3 additional datasets. We rank first on the Medical Segmentation Decathlon (MSD) public leaderboard and achieve state-of-the-art results on Beyond The Cranial Vault (BTCV). Additionally, the Universal Model is computationally more efficient (6x faster) compared with dataset-specific models, generalized better to CT scans from varying sites, and shows stronger transfer learning performance on novel tasks.

eess.IV↗
arXiv2023

Robust Object Modeling for Visual Tracking

Object modeling has become a core part of recent tracking frameworks. Current popular tackers use Transformer attention to extract the template feature separately or interactively with the search region. However, separate template learning lacks communication between the template and search regions, which brings difficulty in extracting discriminative target-oriented features. On the other hand, interactive template learning produces hybrid template features, which may introduce potential distractors to the template via the cluttered search regions. To enjoy the merits of both methods, we propose a robust object modeling framework for visual tracking (ROMTrack), which simultaneously models the inherent template and the hybrid template features. As a result, harmful distractors can be suppressed by combining the inherent features of target objects with search regions' guidance. Target-related features can also be extracted using the hybrid template, thus resulting in a more robust object modeling framework. To further enhance robustness, we present novel variation tokens to depict the ever-changing appearance of target objects. Variation tokens are adaptable to object deformation and appearance variations, which can boost overall performance with negligible computation. Experiments show that our ROMTrack sets a new state-of-the-art on multiple benchmarks.

cs.CV↗
arXiv2023

Absorption of Carbon Dioxide in Kerogen Nanopores: A Mechanism Study using the Molecular Dynamics Monte Carlo Method

Carbon capture and storage (CCS) technology has been applied successfully in recent decades to reduce carbon emissions and alleviate global warming. In this regard, shale reservoirs present tremendous potential for carbon dioxide (CO2) sequestration as they have a large number of nanopores. Molecular dynamics (MD) and MD-Monte Carlo (MDMC) methods were employed in this work to study the absorption behavior of CO2 in shale organic porous media. The MDMC method is used to analyze the spatial states of CO2, and the results are in good agreement with MD results, and it also performs well in the acceleration compared to the classical MD. With regard to the kerogen matrix, its properties, such as the pore size distribution (PSD), pore volume, and surface area, are determined to describe its different compression states and the effects of CO2 absorption on it. The potential energy distribution and potential of mean force are analyzed to verify the spatial distribution of CO2 molecules. The heterogeneity of the pore structure resulted in heterogeneous distributions of CO2 molecules in kerogen porous media. Moreover, strong compression of the matrix reduces the number of large pores, and the PSD is mainly between 0 and 15 Angstrom. Despite the high interaction force of the kerogen matrix, the high-potential-energy region induced by the kerogen skeleton also contributes to the formation of low-energy regions that encourage the entry of CO2. An increase in temperature facilitates the absorption process, allowing CO2 molecules to enter the isolated pores with stronger thermal motion, thereby increasing the storage capacity for CO2. However, the development of geothermal energy may not be suitable for CO2 sequestration.

physics.chem-ph↗
arXiv2023

Flashlight Search Medial Axis: A Pixel-Free Pore-Network Extraction Algorithm

Pore-network models (PNMs) have become an important tool in the study of fluid flow in porous media over the last few decades, and the accuracy of their results highly depends on the extraction of pore networks. Traditional methods of pore-network extraction are based on pixels and require images with high quality. Here, a pixel-free method called the flashlight search medial axis (FSMA) algorithm is proposed for pore-network extraction in a continuous space. The search domain in a two-dimensional space is a line, whereas a surface domain is searched in a three-dimensional scenario. Thus, the FSMA algorithm follows the dimensionality reduction idea; the medial axis can be identified using only a few points instead of calculating every point in the void space. In this way, computational complexity of this method is greatly reduced compared to that of traditional pixel-based extraction methods, thus enabling large-scale pore-network extraction. Based on cases featuring two- and three-dimensional porous media, the FSMA algorithm performs well regardless of the topological structure of the pore network or the positions of the pore and throat centers. This algorithm can also be used to examine both closed- and open-boundary cases. Finally, the FSMA algorithm can search dead-end pores, which is of great significance in the study of multiphase flow in porous media.

cs.CV↗
arXiv2023

Matrix Spherical Functions for $(\mathrm{SU}(n+m),\mathrm{S}(\mathrm{U}(n)\times \mathrm{U}(m)))$: Two Specific Classes

We consider the matrix spherical function related to the compact symmetric pair $(G,K)=(\mathrm{SU}(n+m),\mathrm{S}(\mathrm{U}(n)\times\mathrm{U}(m)))$. The irreducible $K$ representations $(π,V)$ in the ${\rm U}(n)$ part are considered and the induced representation $\mathrm{Ind}_K^Gπ$ splits multiplicity free. In this case, the irreducible $K$ representations in the ${\rm U}(n)$ part are studied. The corresponding spherical functions can be approximated in terms of the simpler matrix-valued functions. We can determine the explicit spherical functions using the action of a differential operator. We consider several cases of irreducible $K$ representations and the orthogonality relations are also described.

math.RT↗
arXiv2023

Lightweight Super-Resolution Head for Human Pose Estimation

Heatmap-based methods have become the mainstream method for pose estimation due to their superior performance. However, heatmap-based approaches suffer from significant quantization errors with downscale heatmaps, which result in limited performance and the detrimental effects of intermediate supervision. Previous heatmap-based methods relied heavily on additional post-processing to mitigate quantization errors. Some heatmap-based approaches improve the resolution of feature maps by using multiple costly upsampling layers to improve localization precision. To solve the above issues, we creatively view the backbone network as a degradation process and thus reformulate the heatmap prediction as a Super-Resolution (SR) task. We first propose the SR head, which predicts heatmaps with a spatial resolution higher than the input feature maps (or even consistent with the input image) by super-resolution, to effectively reduce the quantization error and the dependence on further post-processing. Besides, we propose SRPose to gradually recover the HR heatmaps from LR heatmaps and degraded features in a coarse-to-fine manner. To reduce the training difficulty of HR heatmaps, SRPose applies SR heads to supervise the intermediate features in each stage. In addition, the SR head is a lightweight and generic head that applies to top-down and bottom-up methods. Extensive experiments on the COCO, MPII, and CrowdPose datasets show that SRPose outperforms the corresponding heatmap-based approaches. The code and models are available at https://github.com/haonanwang0522/SRPose.

cs.CV↗
arXiv2023

Energy transfer and radiation in Hamiltonian nonlinear Klein-Gordon equations: general case

In this paper, we consider Klein-Gordon equations with cubic nonlinearity in three spatial dimensions, which are Hamiltonian perturbations of the linear one with potential. It is assumed that the corresponding Klein-Gordon operator $B = \sqrt{-Δ+ V(x) + m^2} $ admits an arbitrary number of possibly degenerate eigenvalues in $(0, m)$, and hence the unperturbed linear equation has multiple time-periodic solutions known as bound states. In \cite{SW1999}, Soffer and Weinstein discovered a mechanism called Fermi's Golden Rule for this nonlinear system in the case of one simple but relatively large eigenvalue $Ω\in (\frac{m}{3}, m)$, by which energy is transferred from discrete to continuum modes and the solution still decays in time. In particular, the exact energy transfer rate is given. In \cite{LLY22}, we solved the general one simple eigenvalue case. In this paper, we solve this problem in full generality: multiple and simple or degenerate eigenvalues in $(0, m)$. The proof is based on a kind of pseudo-one-dimensional cancellation structure in each eigenspace, a renormalized damping mechanism, and an enhanced damping effect. It also relies on a refined Birkhoff normal form transformation and an accurate generalized Fermi's Golden Rule over those of Bambusi--Cuccagna \cite{BC}.

math.AP↗
arXiv2023

Theoretically Guaranteed Policy Improvement Distilled from Model-Based Planning

Model-based reinforcement learning (RL) has demonstrated remarkable successes on a range of continuous control tasks due to its high sample efficiency. To save the computation cost of conducting planning online, recent practices tend to distill optimized action sequences into an RL policy during the training phase. Although the distillation can incorporate both the foresight of planning and the exploration ability of RL policies, the theoretical understanding of these methods is yet unclear. In this paper, we extend the policy improvement step of Soft Actor-Critic (SAC) by developing an approach to distill from model-based planning to the policy. We then demonstrate that such an approach of policy improvement has a theoretical guarantee of monotonic improvement and convergence to the maximum value defined in SAC. We discuss effective design choices and implement our theory as a practical algorithm -- Model-based Planning Distilled to Policy (MPDP) -- that updates the policy jointly over multiple future time steps. Extensive experiments show that MPDP achieves better sample efficiency and asymptotic performance than both model-free and model-based planning algorithms on six continuous control benchmark tasks in MuJoCo.

cs.AI↗
arXiv2023

Caustic effects on the high-order harmonic generation in graphene

We employ the two-band density-matrix equations and time-dependent density functional theory to calculate high-order harmonic generation (HOHG) in graphene under a femtosecond laser irradiation. Our investigation uncovers a striking harmonic enhancement structure (HES) within a specific energy range of the HOHG spectrum. In this regime, we find the convergence of multiple interband electron-hole recombination trajectories, leading to the zero determinant of the Hessian matrix of the semiclassical action. This trajectory convergence exhibits the characteristics akin to the focusing behavior of light rays, commonly known as caustic effects. In contrast to atom situation, where caustic effects are confined to a narrow energy regime around the HOHG cut-off energy and the enhancement due to caustic trajectory convergence is less apparent, the two-dimensional nature of graphene results in a broad energy region for HOHG enhancement that the caustic trajectories can even dominate the entire interband harmonic generation regime. The magnitude of enhancement is significant and can be estimated to be on the order of $\sim N^{2/3}$, with $N$ representing the harmonic order, according to the catastrophe theory. These mechanisms have broad applicability and hold significant implications for other two-dimensional materials, as well as bulk materials, providing crucial insights into the understanding of HOHG phenomena in diverse material systems.

physics.atom-ph↗
arXiv2023

Auxiliary-Tasks Learning for Physics-Informed Neural Network-Based Partial Differential Equations Solving

Physics-informed neural networks (PINNs) have emerged as promising surrogate modes for solving partial differential equations (PDEs). Their effectiveness lies in the ability to capture solution-related features through neural networks. However, original PINNs often suffer from bottlenecks, such as low accuracy and non-convergence, limiting their applicability in complex physical contexts. To alleviate these issues, we proposed auxiliary-task learning-based physics-informed neural networks (ATL-PINNs), which provide four different auxiliary-task learning modes and investigate their performance compared with original PINNs. We also employ the gradient cosine similarity algorithm to integrate auxiliary problem loss with the primary problem loss in ATL-PINNs, which aims to enhance the effectiveness of the auxiliary-task learning modes. To the best of our knowledge, this is the first study to introduce auxiliary-task learning modes in the context of physics-informed learning. We conduct experiments on three PDE problems across different fields and scenarios. Our findings demonstrate that the proposed auxiliary-task learning modes can significantly improve solution accuracy, achieving a maximum performance boost of 96.62% (averaging 28.23%) compared to the original single-task PINNs. The code and dataset are open source at https://github.com/junjun-yan/ATL-PINN.

cs.LG↗
arXiv2023

Machine Learning Study of the Extended Drug-target Interaction Network informed by Pain Related Voltage-Gated Sodium Channels

Pain is a significant global health issue, and the current treatment options for pain management have limitations in terms of effectiveness, side effects, and potential for addiction. There is a pressing need for improved pain treatments and the development of new drugs. Voltage-gated sodium channels, particularly Nav1.3, Nav1.7, Nav1.8, and Nav1.9, play a crucial role in neuronal excitability and are predominantly expressed in the peripheral nervous system. Targeting these channels may provide a means to treat pain while minimizing central and cardiac adverse effects. In this study, we construct protein-protein interaction (PPI) networks based on pain-related sodium channels and develop a corresponding drug-target interaction (DTI) network to identify potential lead compounds for pain management. To ensure reliable machine learning predictions, we carefully select 111 inhibitor datasets from a pool of over 1,000 targets in the PPI network. We employ three distinct machine learning algorithms combined with advanced natural language processing (NLP)-based embeddings, specifically pre-trained transformer and autoencoder representations. Through a systematic screening process, we evaluate the side effects and repurposing potential of over 150,000 drug candidates targeting Nav1.7 and Nav1.8 sodium channels. Additionally, we assess the ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties of these candidates to identify leads with near-optimal characteristics. Our strategy provides an innovative platform for the pharmacological development of pain treatments, offering the potential for improved efficacy and reduced side effects.

q-bio.BM↗
arXiv2023

Quantum criticality at the boundary of the non-Hermitian regime of a Floquet system

We investigate both analytically and numerically the dynamics of quantum scrambling, characterized by the out-of-time ordered correlators (OTOCs), in a non-Hermitian quantum kicked rotor subject to quantum resonance conditions. Analytical expressions for OTOCs as a function of time are obtained, demonstrating a sudden transition from the linear growth to quadratic growth when the non-Hermitian parameter decays to zero. At this critical point, the rates of the linear growth are found to diverge to infinity, indicating the existence of quantum criticality at the boundary of the non-Hermitian regime. The underlying mechanism of this quantum criticality is uncovered, and possible applications in quantum metrology are discussed.

quant-ph↗
arXiv2023

CC-FedAvg: Computationally Customized Federated Averaging

Federated learning (FL) is an emerging paradigm to train model with distributed data from numerous Internet of Things (IoT) devices. It inherently assumes a uniform capacity among participants. However, due to different conditions such as differing energy budgets or executing parallel unrelated tasks, participants have diverse computational resources in practice. Participants with insufficient computation budgets must plan for the use of restricted computational resources appropriately, otherwise they would be unable to complete the entire training procedure, resulting in model performance decline. To address this issue, we propose a strategy for estimating local models without computationally intensive iterations. Based on it, we propose Computationally Customized Federated Averaging (CC-FedAvg), which allows participants to determine whether to perform traditional local training or model estimation in each round based on their current computational budgets. Both theoretical analysis and exhaustive experiments indicate that CC-FedAvg has the same convergence rate and comparable performance as FedAvg without resource constraints. Furthermore, CC-FedAvg can be viewed as a computation-efficient version of FedAvg that retains model performance while considerably lowering computation overhead.

cs.LG↗
arXiv2023

TranssionADD: A multi-frame reinforcement based sequence tagging model for audio deepfake detection

Thanks to recent advancements in end-to-end speech modeling technology, it has become increasingly feasible to imitate and clone a user`s voice. This leads to a significant challenge in differentiating between authentic and fabricated audio segments. To address the issue of user voice abuse and misuse, the second Audio Deepfake Detection Challenge (ADD 2023) aims to detect and analyze deepfake speech utterances. Specifically, Track 2, named the Manipulation Region Location (RL), aims to pinpoint the location of manipulated regions in audio, which can be present in both real and generated audio segments. We propose our novel TranssionADD system as a solution to the challenging problem of model robustness and audio segment outliers in the trace competition. Our system provides three unique contributions: 1) we adapt sequence tagging task for audio deepfake detection; 2) we improve model generalization by various data augmentation techniques; 3) we incorporate multi-frame detection (MFD) module to overcome limited representation provided by a single frame and use isolated-frame penalty (IFP) loss to handle outliers in segments. Our best submission achieved 2nd place in Track 2, demonstrating the effectiveness and robustness of our proposed system.

cs.SD↗
arXiv2023

DVFO: Learning-Based DVFS for Energy-Efficient Edge-Cloud Collaborative Inference

Due to limited resources on edge and different characteristics of deep neural network (DNN) models, it is a big challenge to optimize DNN inference performance in terms of energy consumption and end-to-end latency on edge devices. In addition to the dynamic voltage frequency scaling (DVFS) technique, the edge-cloud architecture provides a collaborative approach for efficient DNN inference. However, current edge-cloud collaborative inference methods have not optimized various compute resources on edge devices. Thus, we propose DVFO, a novel DVFS-enabled edge-cloud collaborative inference framework, which co-optimizes DVFS and offloading parameters via deep reinforcement learning (DRL). Specifically, DVFO automatically co-optimizes 1) the CPU, GPU and memory frequencies of edge devices, and 2) the feature maps to be offloaded to cloud servers. In addition, it leverages a thinking-while-moving concurrent mechanism to accelerate the DRL learning process, and a spatial-channel attention mechanism to extract DNN feature maps of secondary importance for workload offloading. This approach improves inference performance for different DNN models under various edge-cloud network conditions. Extensive evaluations using two datasets and six widely-deployed DNN models on three heterogeneous edge devices show that DVFO significantly reduces the energy consumption by 33% on average, compared to state-of-the-art schemes. Moreover, DVFO achieves up to 28.6%-59.1% end-to-end latency reduction, while maintaining accuracy within 1% loss on average.

cs.LG↗
arXiv2023

Phonon-phonon interactions in the polarizarion dependence of Raman scattering

We have found that the polarization dependence of Raman scattering in organic crystals at finite temperatures can only be described by a fourth-rank formalism. This generalization of the second-rank Raman tensor $\mathcal{R}$ stems from the effect of off-diagonal components in the crystal self-energy on the light scattering mechanism. We thus establish a novel manifestation of phonon-phonon interaction in inelastic light scattering, markedly separate from the better-known phonon lifetime.

cond-mat.mtrl-sci↗
arXiv2023

ST-PINN: A Self-Training Physics-Informed Neural Network for Partial Differential Equations

Partial differential equations (PDEs) are an essential computational kernel in physics and engineering. With the advance of deep learning, physics-informed neural networks (PINNs), as a mesh-free method, have shown great potential for fast PDE solving in various applications. To address the issue of low accuracy and convergence problems of existing PINNs, we propose a self-training physics-informed neural network, ST-PINN. Specifically, ST-PINN introduces a pseudo label based self-learning algorithm during training. It employs governing equation as the pseudo-labeled evaluation index and selects the highest confidence examples from the sample points to attach the pseudo labels. To our best knowledge, we are the first to incorporate a self-training mechanism into physics-informed learning. We conduct experiments on five PDE problems in different fields and scenarios. The results demonstrate that the proposed method allows the network to learn more physical information and benefit convergence. The ST-PINN outperforms existing physics-informed neural network methods and improves the accuracy by a factor of 1.33x-2.54x. The code of ST-PINN is available at GitHub: https://github.com/junjun-yan/ST-PINN.

cs.LG↗
arXiv2023

Variational Bayesian Inference for Bipartite Mixed-membership Stochastic Block Model with Applications to Collaborative Filtering

Motivated by the connections between collaborative filtering and network clustering, we consider a network-based approach to improving rating prediction in recommender systems. We propose a novel Bipartite Mixed-Membership Stochastic Block Model ($\mathrm{BM}^2$) with a conjugate prior from the exponential family. We derive the analytical expression of the model and introduce a variational Bayesian expectation-maximization algorithm, which is computationally feasible for approximating the untractable posterior distribution. We carry out extensive simulations to show that $\mathrm{BM}^2$ provides more accurate inference than standard SBM with the emergence of outliers. Finally, we apply the proposed model to a MovieLens dataset and find that it outperforms other competing methods for collaborative filtering.

stat.ME↗