arXiv subjects

Jie Li

Publications and source records attributed to Jie Li.

At least 469 records · Page 26Linked to original sources

arXiv2022

Learning to Evolve Structural Ensembles of Unfolded and Disordered Proteins Using Experimental Solution Data

We have developed a Generative Recurrent Neural Networks (GRNN) that learns the probability of the next residue torsions $X_{i+1}=\ [ϕ_{i+1},ψ_{i+1},ω_{i+1}, χ_{i+1}]$ from the previous residue in the sequence $X_i$ to generate new IDP conformations. In addition, we couple the GRNN with a Bayesian model, X-EISD, in a reinforcement learning step that biases the probability distributions of torsions to take advantage of experimental data types such as J-couplingss, NOEs and PREs. We show that updating the generative model parameters according to the reward feedback on the basis of the agreement between structures and data improves upon existing approaches that simply reweight static structural pools for disordered proteins. Instead the GRNN "DynamICE" model learns to physically change the conformations of the underlying pool to those that better agree with experiment.

physics.bio-ph↗

arXiv2022

TransFA: Transformer-based Representation for Face Attribute Evaluation

Face attribute evaluation plays an important role in video surveillance and face analysis. Although methods based on convolution neural networks have made great progress, they inevitably only deal with one local neighborhood with convolutions at a time. Besides, existing methods mostly regard face attribute evaluation as the individual multi-label classification task, ignoring the inherent relationship between semantic attributes and face identity information. In this paper, we propose a novel \textbf{trans}former-based representation for \textbf{f}ace \textbf{a}ttribute evaluation method (\textbf{TransFA}), which could effectively enhance the attribute discriminative representation learning in the context of attention mechanism. The multiple branches transformer is employed to explore the inter-correlation between different attributes in similar semantic regions for attribute feature learning. Specially, the hierarchical identity-constraint attribute loss is designed to train the end-to-end architecture, which could further integrate face identity discriminative information to boost performance. Experimental results on multiple face attribute benchmarks demonstrate that the proposed TransFA achieves superior performances compared with state-of-the-art methods.

cs.CV↗

arXiv2022

SpOT: Spatiotemporal Modeling for 3D Object Tracking

3D multi-object tracking aims to uniquely and consistently identify all mobile entities through time. Despite the rich spatiotemporal information available in this setting, current 3D tracking methods primarily rely on abstracted information and limited history, e.g. single-frame object bounding boxes. In this work, we develop a holistic representation of traffic scenes that leverages both spatial and temporal information of the actors in the scene. Specifically, we reformulate tracking as a spatiotemporal problem by representing tracked objects as sequences of time-stamped points and bounding boxes over a long temporal history. At each timestamp, we improve the location and motion estimates of our tracked objects through learned refinement over the full sequence of object history. By considering time and space jointly, our representation naturally encodes fundamental physical priors such as object permanence and consistency across time. Our spatiotemporal tracking framework achieves state-of-the-art performance on the Waymo and nuScenes benchmarks.

cs.CV↗

arXiv2022

Fully Attentional Network for Semantic Segmentation

Recent non-local self-attention methods have proven to be effective in capturing long-range dependencies for semantic segmentation. These methods usually form a similarity map of RC*C (by compressing spatial dimensions) or RHW*HW (by compressing channels) to describe the feature relations along either channel or spatial dimensions, where C is the number of channels, H and W are the spatial dimensions of the input feature map. However, such practices tend to condense feature dependencies along the other dimensions,hence causing attention missing, which might lead to inferior results for small/thin categories or inconsistent segmentation inside large objects. To address this problem, we propose anew approach, namely Fully Attentional Network (FLANet),to encode both spatial and channel attentions in a single similarity map while maintaining high computational efficiency. Specifically, for each channel map, our FLANet can harvest feature responses from all other channel maps, and the associated spatial positions as well, through a novel fully attentional module. Our new method has achieved state-of-the-art performance on three challenging semantic segmentation datasets,i.e., 83.6%, 46.99%, and 88.5% on the Cityscapes test set,the ADE20K validation set, and the PASCAL VOC test set,respectively.

cs.CV↗

arXiv2022

The quasi-Keplerian motion in the regular Bardeen spacetime

The second post-Newtonian solution for the quasi-Keplerian motion of a test particle in the gravitational field of regular Bardeen black hole is derived. The solution is formulated in terms of the test particle's orbital energy and angular momentum, as well as the mass and magnetic charge of the Bardeen black hole. The leading effects of the magnetic charge on the test particle's orbit and motion including perihelion precession are displayed explicitly. In particular, it is shown that to the second post-Newtonian order the magnetic charge does not affect the test particle's orbital period.

gr-qc↗

arXiv2022

Digital-twin-enhanced metal tube bending forming real-time prediction method based on Multi-source-input MTL

As one of the most widely used metal tube bending methods, the rotary draw bending (RDB) process enables reliable and high-precision metal tube bending forming (MTBF). The forming accuracy is seriously affected by the springback and other potential forming defects, of which the mechanism analysis is difficult to deal with. At the same time, the existing methods are mainly conducted in offline space, ignoring the real-time information in the physical world, which is unreliable and inefficient. To address this issue, a digital-twin-enhanced (DT-enhanced) metal tube bending forming real-time prediction method based on multi-source-input multi-task learning (MTL) is proposed. The new method can achieve comprehensive MTBF real-time prediction. By sharing the common feature of the multi-close domain and adopting group regularization strategy on feature sharing and accepting layers, the accuracy and efficiency of the multi-source-input MTL can be guaranteed. Enhanced by DT, the physical real-time deformation data is aligned in the image dimension by an improved Grammy Angle Field (GAF) conversion, realizing the reflection of the actual processing. Different from the traditional offline prediction methods, the new method integrates the virtual and physical data to achieve a more efficient and accurate real-time prediction result. and the DT mapping connection between virtual and physical systems can be achieved. To exclude the effects of equipment errors, the effectiveness of the proposed method is verified on the physical experiment-verified FE simulation scenarios. At the same time, the common pre-training networks are compared with the proposed method. The results show that the proposed DT-enhanced prediction method is more accurate and efficient.

cs.LG↗

arXiv2022

Magnon squeezing enhanced ground-state cooling in cavity magnomechanics

Cavity magnomechanics has recently become a new platform for studying macroscopic quantum phenomena. The magnetostriction induced vibration mode of a large-size ferromagnet or ferrimagnet reaching its ground state represents a genuine macroscopic quantum state. Here we study the ground-state cooling of the mechanical vibration mode in a cavity magnomechanical system, and focus on the role of magnon squeezing in improving the cooling efficiency. The magnon squeezing is obtained by exploiting the magnon self-Kerr nonlinearity. We find that the magnon squeezing can significantly and even completely suppress the magnomechanical Stokes scattering. It thus becomes particularly useful in realizing ground-state cooling in the unresolved-sideband regime, where the conventional sideband cooling protocols become inefficient. We also find that the coupling to the microwave cavity plays only an adverse effect in mechanical cooling. This makes essentially the two-mode magnomechanical system (without involving the microwave cavity) a preferred system for cooling the mechanical motion, in which the magnon mode is established by a uniform bias magnetic field and a microwave drive field.

quant-ph↗

arXiv2022

Spin transfer from dark matter to gas during halo formation

In the protogalactic density field, diffuse gas and collision-less cold dark matter (DM) are often assumed sufficiently mixed that both components experience identical tidal torques. However, haloes in cosmological simulations consistently end up with a higher specific angular momentum (sAM) in gas, even in simulations without radiative cooling and galaxy formation physics. We refine this result by analysing the spin distributions of gas and DM in $\sim$50,000 well-resolved haloes in a non-radiative cosmological simulation from the SURFS suite. The sAM of the halo gas on average ends up $\sim$40\% above that of the DM. This can be pinned down to an excess AM in the inner halo ($<$50\% virial radius), paralleled by a more coherent rotation pattern in the gas. We uncover the leading driver for this AM difference through a series of control simulations of a collapsing ellipsoidal top-hat, where gas and DM are initially well mixed. These runs reveal that the pressurised inner gas shells collapse more slowly, causing the DM ellipsoid to spin ahead of the gas ellipsoid. The arising torque generally transfers AM from the DM to the gas. The amount of AM transferred via this mode depends on the initial spin, the initial axes ratios and the collapse factor. These quantities can be combined in a single dimensionless parameter, which robustly predicts the AM transfer of the ellipsoidal collapse. This simplistic model can quantitatively explain the average AM excess of the gas found in the more complex non-radiative cosmological simulation.

astro-ph.GA↗

arXiv2022

Object Permanence Emerges in a Random Walk along Memory

This paper proposes a self-supervised objective for learning representations that localize objects under occlusion - a property known as object permanence. A central question is the choice of learning signal in cases of total occlusion. Rather than directly supervising the locations of invisible objects, we propose a self-supervised objective that requires neither human annotation, nor assumptions about object dynamics. We show that object permanence can emerge by optimizing for temporal coherence of memory: we fit a Markov walk along a space-time graph of memories, where the states in each time step are non-Markovian features from a sequence encoder. This leads to a memory representation that stores occluded objects and predicts their motion, to better localize them. The resulting model outperforms existing approaches on several datasets of increasing complexity and realism, despite requiring minimal supervision, and hence being broadly applicable.

cs.CV↗

arXiv2022

Predicting Axillary Lymph Node Metastasis in Early Breast Cancer Using Deep Learning on Primary Tumor Biopsy Slides

Objectives: To develop and validate a deep learning (DL)-based primary tumor biopsy signature for predicting axillary lymph node (ALN) metastasis preoperatively in early breast cancer (EBC) patients with clinically negative ALN. Methods: A total of 1,058 EBC patients with pathologically confirmed ALN status were enrolled from May 2010 to August 2020. A DL core-needle biopsy (DL-CNB) model was built on the attention-based multiple instance-learning (AMIL) framework to predict ALN status utilizing the DL features, which were extracted from the cancer areas of digitized whole-slide images (WSIs) of breast CNB specimens annotated by two pathologists. Accuracy, sensitivity, specificity, receiver operating characteristic (ROC) curves, and areas under the ROC curve (AUCs) were analyzed to evaluate our model. Results: The best-performing DL-CNB model with VGG16_BN as the feature extractor achieved an AUC of 0.816 (95% confidence interval (CI): 0.758, 0.865) in predicting positive ALN metastasis in the independent test cohort. Furthermore, our model incorporating the clinical data, which was called DL-CNB+C, yielded the best accuracy of 0.831 (95%CI: 0.775, 0.878), especially for patients younger than 50 years (AUC: 0.918, 95%CI: 0.825, 0.971). The interpretation of DL-CNB model showed that the top signatures most predictive of ALN metastasis were characterized by the nucleus features including density ($p$ = 0.015), circumference ($p$ = 0.009), circularity ($p$ = 0.010), and orientation ($p$ = 0.012). Conclusion: Our study provides a novel DL-based biomarker on primary tumor CNB slides to predict the metastatic status of ALN preoperatively for patients with EBC. The codes and dataset are available at https://github.com/bupt-ai-cz/BALNMP

eess.IV↗

arXiv2022

Ultra-narrowband terahertz circular dichroism driven by planar metasurface supporting chiral quasi bound states in continuum

Terahertz (THz) chirality pursues customizable manipulation from narrowband to broadband. While conventional THz chirality is restricted by non-negligible linewidth and unable to handle narrowband well. Recently, the concept "quasi bound states in continuum" (quasi-BIC) is introduced to optics resonance system whose the quality factor can be extremely high with the ultra-low radiative loss, thus providing a conceptual feasibility for wave control with ultra-narrow linewidth. Herein, we construct quasi-BIC in a planar all-silicon THz metasurface with in-plane C2 and mirror symmetries breaking. Such system not only exposes the symmetry-protected BIC, but also exposes the parameter-tuned BIC assigned to single resonance type. An extremely narrow linewidth (below 0.06 GHz) with high quality factor (104 level) is obtained at quasi-BIC frequency, which achieves the ultra-narrowband THz chirality.

physics.optics↗

arXiv2022

BGP-Multipath Routing in the Internet

BGP-Multipath (BGP-M) is a multipath routing technique for load balancing. Distinct from other techniques deployed at a router inside an Autonomous System (AS), BGP-M is deployed at a border router that has installed multiple inter-domain border links to a neighbour AS. It uses the equal-cost multi-path (ECMP) function of a border router to share traffic to a destination prefix on different border links. Despite recent research interests in multipath routing, there is little study on BGP-M. Here we provide the first measurement and a comprehensive analysis of BGP-M routing in the Internet. We extracted information on BGP-M from query data collected from Looking Glass (LG) servers. We revealed that BGP-M has already been extensively deployed and used in the Internet. A particular example is Hurricane Electric (AS6939), a Tier-1 network operator, which has implemented >1,000 cases of BGP-M at 69 of its border routers to prefixes in 611 of its neighbour ASes, including many hyper-giant ASes and large content providers, on both IPv4 and IPv6 Internet. We examined the distribution and operation of BGP-M. We also ran traceroute using RIPE Atlas to infer the routing paths, the schemes of traffic allocation, and the delay on border links. This study provided the state-of-the-art knowledge on BGP-M with novel insights into the unique features and the distinct advantages of BGP-M as an effective and readily available technique for load balancing.

cs.NI↗

arXiv2022

Autonomous Smart Grid Fault Detection

Smart grid plays a crucial role for the smart society and the upcoming carbon neutral society. Achieving autonomous smart grid fault detection is critical for smart grid system state awareness, maintenance and operation. This paper focuses on fault monitoring in smart grid and discusses the inherent technical challenges and solutions. In particular, we first present the basic principles of smart grid fault detection. Then, we explain the new requirements for autonomous smart grid fault detection, the technical challenges and their possible solutions. A case study is introduced, as a preliminary study for autonomous smart grid fault detection. In addition, we highlight relevant directions for future research.

cs.DC↗

arXiv2022

On Understanding and Mitigating the Dimensional Collapse of Graph Contrastive Learning: a Non-Maximum Removal Approach

Graph Contrastive Learning (GCL) has shown promising performance in graph representation learning (GRL) without the supervision of manual annotations. GCL can generate graph-level embeddings by maximizing the Mutual Information (MI) between different augmented views of the same graph (positive pairs). However, the GCL is limited by dimensional collapse, i.e., embedding vectors only occupy a low-dimensional subspace. In this paper, we show that the smoothing effect of the graph pooling and the implicit regularization of the graph convolution are two causes of the dimensional collapse in GCL. To mitigate the above issue, we propose a non-maximum removal graph contrastive learning approach (nmrGCL), which removes "prominent'' dimensions (i.e., contribute most in similarity measurement) for positive pair in the pre-text task. Comprehensive experiments on various benchmark datasets are conducted to demonstrate the effectiveness of nmrGCL, and the results show that our model outperforms the state-of-the-art methods. Source code will be made publicly available.

cs.LG↗

arXiv2022

Steady entangled-state generation via cross-Kerr effect in a ferrimagnetic crystal

For solid-state spin systems, the collective spin motion in a single crystal embodies multiple magnetostatic modes. Recently, it was found that the cross-Kerr interaction between the higher-order magnetostatic mode and the Kittel mode introduces a new operable degree of freedom. In this work, we propose a scheme to entangle two magnon modes via the cross-Kerr nonlinearity when the bias field is inhomogeneous and the system is driven. Quantum entanglement persists at the steady state, as demonstrated by numerical results using experimentally feasible parameters. Furthermore, we also demonstrate that entangled states can survive better in the system where self-Kerr and cross-Kerr nonlinearities coexist. Our work provides insights and guidance for designing experiments to observe entanglement between different degrees of freedom within a single ferrimagnetic crystal. Additionally, it may stimulate potential applications in quantum information processing using spintronic devices.

quant-ph↗

arXiv2022

Double crossed biproducts and related structures

Let $H$ be a bialgebra. Let $σ: H\otimes H\to A$ be a linear map, where $A$ is a left $H$-comodule coalgebra, and an algebra with a left $H$-weak action $\triangleright$. Let $τ: H\otimes H\to B$ be a linear map, where $B$ is a right $H$-comodule coalgebra, and an algebra with a right $H$-weak action $\triangleleft$. In this paper, we improve the necessary conditions for the two-sided crossed product algebra $A\#^σ H~{^τ\#} B$ and the two-sided smash coproduct coalgebra $A\times H\times B$ to form a bialgebra (called double crossed biproduct) such that the condition $b_{[1]}\triangleright a_0\otimes b_{[0]}\triangleleft a_{-1}=a\otimes b$ in Majid's double biproduct (or double-bosonization) is one of the necessary conditions. On the other hand, we provide a more general two-sided crossed product algebra structure via Brzezński's crossed product and give some applications.

math.RA↗

arXiv2022

Learning Correlations between Internal Coordinates to improve 3D Cartesian Coordinates for Proteins

We consider a generic representation problem of internal coordinates (bond lengths, valence angles, and dihedral angles) and their transformation to 3-dimensional Cartesian coordinates of a biomolecule. We show that the internal-to-Cartesian process relies on correctly predicting chemically subtle correlations among the internal coordinates themselves, and learning these correlations increases the fidelity of the Cartesian representation. This general problem has been solved with machine learning for proteins, but with appropriately formulated data is extensible to any type of chain biomolecule including RNA, DNA, and lipids. We show that the internal-to-Cartesian process relies on correctly predicting chemically subtle correlations among the internal coordinates themselves, and learning these correlations increases the fidelity of the Cartesian representation. We developed a machine learning algorithm, Int2Cart, to predict bond lengths and bond angles from backbone torsion angles and residue types of a protein, and allows reconstruction of protein structures better than using fixed bond lengths and bond angles, or a static library method that relies on backbone torsion angles and residue types on a single residue. The Int2Cart algorithm has been implemented as an individual python package at https://github.com/THGLab/int2cart.

physics.bio-ph↗

arXiv2022

A Smart Contract based Crowdfunding Mechanism for Hierarchical Federated Learning

Hierarchical Federated Learning (HFL) is introduced as a promising technique that allows model owners to fully exploit computational resources and bandwidth resources to train the global model. However, due to the high training cost, a single model owner may not be able to deploy HFL. To address this issue, we develop a smart contract based trust crowdfunding mechanism for HFL, which enables multiple model owners to obtain a crowdfunding model with high social utility for multiple crowdfunding participants. To ensure the authenticity of the crowdfunding mechanism, we implemented the Vickey-Clark-Croves (VCG) mechanism to encourage all crowdfunding participants and clients to provide realistic bids and offers. At the same time, in order to ensure guaranteed trustworthiness of crowdfunding and automatic distribution of funds, we develop and implement a smart contract to record the crowdfunding process and training results in the blockchain. We prove that the proposed scheme satisfies the budget balance and participant constraint. Finally, we implement a prototype of this smart contract on an Ethereoum private chain and evaluate the proposed VCG mechanism. The experimental results demonstrate that the proposed scheme can effectively improve social utility while ensuring the authenticity and trustworthiness of the crowdfunding process.

cs.DC↗