arXiv subjects

Jing Liu

Publications and source records attributed to Jing Liu.

At least 541 records · Page 30Linked to original sources

arXiv2021

Micro-supervised Disturbance Learning: A Perspective of Representation Probability Distribution

The instability is shown in the existing methods of representation learning based on Euclidean distance under a broad set of conditions. Furthermore, the scarcity and high cost of labels prompt us to explore more expressive representation learning methods which depends on the labels as few as possible. To address these issues, the small-perturbation ideology is firstly introduced on the representation learning model based on the representation probability distribution. The positive small-perturbation information (SPI) which only depend on two labels of each cluster is used to stimulate the representation probability distribution and then two variant models are proposed to fine-tune the expected representation distribution of RBM, namely, Micro-supervised Disturbance GRBM (Micro-DGRBM) and Micro-supervised Disturbance RBM (Micro-DRBM) models. The Kullback-Leibler (KL) divergence of SPI is minimized in the same cluster to promote the representation probability distributions to become more similar in Contrastive Divergence(CD) learning. In contrast, the KL divergence of SPI is maximized in the different clusters to enforce the representation probability distributions to become more dissimilar in CD learning. To explore the representation learning capability under the continuous stimulation of the SPI, we present a deep Micro-supervised Disturbance Learning (Micro-DL) framework based on the Micro-DGRBM and Micro-DRBM models and compare it with a similar deep structure which has not any external stimulation. Experimental results demonstrate that the proposed deep Micro-DL architecture shows better performance in comparison to the baseline method, the most related shallow models and deep frameworks for clustering.

cs.LG↗

arXiv2021

Radiation-hardened and Repairable MoS$_2$ Field Effect Devices with Polymer Solid Electrolyte Gates

As human activities expand into naturally or man-made radiation-prone environment, the need for radiation-hardened (Rad-Hard) electronic hardware surged. The state-of-the-art silicon-based and two-dimensional (2D) materials based Rad-Hard transistors can withstand up to 1 Mrad (Si) of total ionization dose (TID), while higher TID tolerance is being heatedly sought after. Here we present few-layer MoS$_2$ Rad-Hard field-effect transistors (FETs) with polymer solid electrolyte (PSE) gate dielectrics. The MoS$_2$ PSE-FETs exhibit a TID tolerance of up to 3.75 Mrad (Si) at a dose rate of 523 rad (Si) s$^{-1}$ and can be repaired with a moderate thermal annealing at 100 $^{\circ}$C for 5 minutes. Combining the excellent intrinsic radiation tolerance and the reparability, the MoS$_2$ PSE-FETs reach a TID tolerance of up to 10 Mrad (Si). Complementary metal-oxide-semiconductor (CMOS)-like MoS$_2$ PSE-inverters have been built and show similar high radiation tolerance. Furthermore, the feasibility of wafer-scale Rad-Hard PSE-inverter array has been demonstrated using chemical vapor deposition (CVD) grown monolayer MoS$_2$. Our studies uncover the potential of 2D materials based PSE devices in future Rad-Hard integrated circuits (ICs).

cond-mat.mtrl-sci↗

arXiv2021

Fluidic Endogenous Magnetism and Magnetic Monopole Clues from Liquid Metal Droplet Machine

Magnetism and magnetic monopole are classical issues in basic physics. Conventional magnets are generally composed of rigid materials with shapes and structures unchangeable which may face challenges sometimes to answer the above questions. Here, from an alternative other than rigid magnet, we disclosed an unconventional way to generate endogenous magnetism and then construct magnetic monopole through tuning liquid metal machine. Through theoretical interpretation and conceptual experiments, we illustrated that when gallium base liquid metal in solution rotates under actuation of an external electric field, it forms an endogenous magnetic field inside which well explains the phenomenon that two such discrete metal droplets could easily fuse together, indicating their reciprocal attraction via N and S poles. Further, we conceived that the self-driving liquid metal motor was also an endogenous magnet owning the electromagnetic homology. When liquid metal in solution swallowed aluminum inside, it formed a spin motor and dynamically variable charge distribution which produced an endogenous magnetic field. This finding explains the phenomena that there often happened reflection collision and attraction fusion between running liquid metal motors which were just caused by the dynamic adjustment of their N and S polarities. Finally, we conceived that such endogenous magnet could lead to magnetic monopole and four technical routes were suggested as: 1. Matching the interior flow field of liquid metal machines; 2. Superposition between external electric effect and magnetic field; 3. Composite construction between magnetic particles and liquid metal motor; 4. Chemical ways such as via galvanic cell reaction. Overall, the fluidic endogenous magnet and the promising magnetic monopole it enabled may lead to unconventional magnetoelectric devices and applications in the near future.

cond-mat.soft↗

arXiv2021

Temporal Memory Attention for Video Semantic Segmentation

Video semantic segmentation requires to utilize the complex temporal relations between frames of the video sequence. Previous works usually exploit accurate optical flow to leverage the temporal relations, which suffer much from heavy computational cost. In this paper, we propose a Temporal Memory Attention Network (TMANet) to adaptively integrate the long-range temporal relations over the video sequence based on the self-attention mechanism without exhaustive optical flow prediction. Specially, we construct a memory using several past frames to store the temporal information of the current frame. We then propose a temporal memory attention module to capture the relation between the current frame and the memory to enhance the representation of the current frame. Our method achieves new state-of-the-art performances on two challenging video semantic segmentation datasets, particularly 80.3% mIoU on Cityscapes and 76.5% mIoU on CamVid with ResNet-50.

cs.CV↗

arXiv2021

Exploiting Spatial-Temporal Semantic Consistency for Video Scene Parsing

Compared with image scene parsing, video scene parsing introduces temporal information, which can effectively improve the consistency and accuracy of prediction. In this paper, we propose a Spatial-Temporal Semantic Consistency method to capture class-exclusive context information. Specifically, we design a spatial-temporal consistency loss to constrain the semantic consistency in spatial and temporal dimensions. In addition, we adopt an pseudo-labeling strategy to enrich the training dataset. We obtain the scores of 59.84% and 58.85% mIoU on development (test part 1) and testing set of VSPW, respectively. And our method wins the 1st place on VSPW challenge at ICCV2021.

cs.CV↗

arXiv2021

Probing the $B+L$ violation process with the observation of cosmic magnetic field

We investigate the $B+L$ violation process by performing three-dimensional lattice simulations in an electroweak theory with first-order phase transition and the electroweak sphaleron decay. The simulation results indicate that the Chern-Simons number changes along with the helical magnetic field production when the sphaleron decay occurs. Our study suggests that, for the electroweak phase transition with nucleation rate being smaller than $\sim \mathcal{O}(10)$, the helical magnetic field with the fractional magnetic helicity $ε_M\leq 0.2$ can be probed by Cherenkov Telescope Array through the intergalactic magnetic field measurements. Based on our numerical results, we suggest a method to probe the baryon asymmetry generation of the Universe, which is a general consequence of the electroweak sphaleron process, through the astronomical observation of the corresponding helical magnetic field.

hep-ph↗

arXiv2021

Thermal conductivity of micro/nano-porous polymers: Prediction models and applications

Micro/nano porous polymeric material is considered a unique industrial material due to its extremely low thermal conductivity, low density, and high surface area. Therefore, it is necessary to establish an accurate thermal conductivity prediction model suiting their applicable conditions and provide a theoretical basis for expanding of their applications. In this work, the development of the calculation model of equivalent thermal conductivity of micro/nano porous polymeric materials in recent years was summarized. Firstly, it reviews the process of establishing the overall equivalent thermal conductivity calculation model for micro/nano porous polymers. Then, the predicted calculation models of thermal conductivity are introduced according to the conductive thermal conductivity models and the radiative thermal conductivity models separately. In addition, the thermal conduction part is divided into the gaseous thermal conductivity model, solid thermal conductivity model and gas-solid coupling model. Finally, it is concluded that, compared with other porous materials, there are few studies on heat transfer of micro/nano porous polymers, especially on the special heat transfer mechanisms such as scale effects at the micro/nanoscale. In particular, the following aspects of porous polymers still need to be further studied: micro scaled thermal radiation, heat transfer characteristics of special morphologies at the nanoscale, heat transfer mechanism and impact factors of micro/nano porous polymers. Such studies would provide a more accurate prediction of thermal conductivity and a wider application in energy conversion and storage systems.

physics.comp-ph↗

arXiv2021

Quantized orbital-chasing liquid metal heterodimers directed by an integrated pilot-wave field

A millimetric bouncing droplet sustained on a vibrating bath becomes a moving wave source (particle) through periodically interacting with the local wave field it generates during the droplet-bath impact. By virtue of such particle-wave duality, the macroscopic hydrodynamic system imitates enigmatic behaviors of the quantum realm. Here we show that it is possible to create an integrated pilot-wave field to better prescribe the droplet trajectories, via amplified bath capillarity. This is demonstrated with a liquid metal droplet-bath system in which the local wave field generated by droplet bouncing is superposed by the global wave field induced by bath meniscus oscillation. The resulting dual pilot-wave configuration enables a class of directional chasing motions of two bound dissimilar droplets (heterodimers) in multilevel hydrodynamic traps (orbits), featuring two quantized regime parameters, namely the interdroplet binding level and the orbit level. We investigate the dynamics of the vibrating liquid metal bath, with its level-split ring wave field and its peculiar vortex field being highlighted. We also rationalize the exotic droplet motions by considering the interdroplet particle-wave interactions mediated by the integrated pilot-wave field. It is revealed that a temporal bouncing phase shift between the two droplets in the heterodimers, due to size mismatch, gives rise to their horizontal propulsion, while their spatial binding regime exclusively determines the collective chasing direction. It is further evidenced that the horizontal in-orbit chasing motion is directly related to vertical droplet bouncing. Our findings unveil the integrated pilot-wave field as a trail towards improved droplet guiding, thereby extending the hydrodynamic particle-wave analogy to optical systems and beyond.

physics.flu-dyn↗

arXiv2021

Magnetic field and gravitational waves from the first-order Phase Transition

We perform the three dimensional lattice simulation of the magnetic field and gravitational wave productions from bubble collisions during the first-order electroweak phase transition. Except that of the gravitational wave, the power-law spectrum of the magnetic field strength is numerically calculated for the first time, which is of a broken power-law spectrum: $B_ξ\propto f^{0.91}$ for low frequency region of $f f_\star$ in the thin-wall limit, with the peak frequency being $f_\star\sim 5$ Hz at the phase transition temperature 100 GeV. When the hydrodynamics is taken into account, the generated magnetic field strength can reach $B_ξ\sim 10^{-7}$G at a correlation length $ξ\sim 10^{-7}$pc, which may seed the large scale magnetic fields. Our study shows that the measurements of cosmic magnetic field strength and gravitational waves are complementary to probe new physics admitting electroweak phase transition.

astro-ph.CO↗

arXiv2021

Scalable Vision Transformers with Hierarchical Pooling

The recently proposed Visual image Transformers (ViT) with pure attention have achieved promising performance on image recognition tasks, such as image classification. However, the routine of the current ViT model is to maintain a full-length patch sequence during inference, which is redundant and lacks hierarchical representation. To this end, we propose a Hierarchical Visual Transformer (HVT) which progressively pools visual tokens to shrink the sequence length and hence reduces the computational cost, analogous to the feature maps downsampling in Convolutional Neural Networks (CNNs). It brings a great benefit that we can increase the model capacity by scaling dimensions of depth/width/resolution/patch size without introducing extra computational complexity due to the reduced sequence length. Moreover, we empirically find that the average pooled visual tokens contain more discriminative information than the single class token. To demonstrate the improved scalability of our HVT, we conduct extensive experiments on the image classification task. With comparable FLOPs, our HVT outperforms the competitive baselines on ImageNet and CIFAR-100 datasets. Code is available at https://github.com/MonashAI/HVT

cs.CV↗

arXiv2021

Double Exotic States Productions in Pion and Kaon Induced Reactions

In the present work, we investigate the production of $P_{c}$ and $Z_c(3900)/Z_{cs}(3985)$ states in the processes of $πp \rightarrow Z_{c}(3900) P_{c} $/$Kp\rightarrow Z_{cs}(3985) P_{c}$ with an effective Lagrangian approach. Our estimations indicates that the cross sections depends on the model parameters and beam energies, in particular, the cross sections for $πp \to Z_c(3900) P_c(4312)$ and $πp \to Z_c(3900) P_c(4440)$ are similar and can reach up to 10 nb, while those for $πp \to Z_c(3900) P_c(4457)$ can be 30 nb. As for $Kp \to Z_{cs}(3985)P_c$, the cross sections are about 2.5 times smaller than those for $πp \to Z_c(3900) P_c$. The discussed processes in the present work, especially $πp \to Z_c(3900) P_c(4457)$, may be accessible by the the high-energy pion and kaon beams in the facilities of J-PARC and COMPASS.

hep-ph↗

arXiv2021

HANet: Hierarchical Alignment Networks for Video-Text Retrieval

Video-text retrieval is an important yet challenging task in vision-language understanding, which aims to learn a joint embedding space where related video and text instances are close to each other. Most current works simply measure the video-text similarity based on video-level and text-level embeddings. However, the neglect of more fine-grained or local information causes the problem of insufficient representation. Some works exploit the local details by disentangling sentences, but overlook the corresponding videos, causing the asymmetry of video-text representation. To address the above limitations, we propose a Hierarchical Alignment Network (HANet) to align different level representations for video-text matching. Specifically, we first decompose video and text into three semantic levels, namely event (video and text), action (motion and verb), and entity (appearance and noun). Based on these, we naturally construct hierarchical representations in the individual-local-global manner, where the individual level focuses on the alignment between frame and word, local level focuses on the alignment between video clip and textual context, and global level focuses on the alignment between the whole video and text. Different level alignments capture fine-to-coarse correlations between video and text, as well as take the advantage of the complementary information among three semantic levels. Besides, our HANet is also richly interpretable by explicitly learning key semantic concepts. Extensive experiments on two public datasets, namely MSR-VTT and VATEX, show the proposed HANet outperforms other state-of-the-art methods, which demonstrates the effectiveness of hierarchical representation and alignment. Our code is publicly available.

cs.CV↗

arXiv2021

Resisting Out-of-Distribution Data Problem in Perturbation of XAI

With the rapid development of eXplainable Artificial Intelligence (XAI), perturbation-based XAI algorithms have become quite popular due to their effectiveness and ease of implementation. The vast majority of perturbation-based XAI techniques face the challenge of Out-of-Distribution (OoD) data -- an artifact of randomly perturbed data becoming inconsistent with the original dataset. OoD data leads to the over-confidence problem in model predictions, making the existing XAI approaches unreliable. To our best knowledge, the OoD data problem in perturbation-based XAI algorithms has not been adequately addressed in the literature. In this work, we address this OoD data problem by designing an additional module quantifying the affinity between the perturbed data and the original dataset distribution, which is integrated into the process of aggregation. Our solution is shown to be compatible with the most popular perturbation-based XAI algorithms, such as RISE, OCCLUSION, and LIME. Experiments have confirmed that our methods demonstrate a significant improvement in general cases using both computational and cognitive metrics. Especially in the case of degradation, our proposed approach demonstrates outstanding performance comparing to baselines. Besides, our solution also resolves a fundamental problem with the faithfulness indicator, a commonly used evaluation metric of XAI algorithms that appears to be sensitive to the OoD issue.

cs.AI↗

arXiv2021

DuReader_robust: A Chinese Dataset Towards Evaluating Robustness and Generalization of Machine Reading Comprehension in Real-World Applications

Machine reading comprehension (MRC) is a crucial task in natural language processing and has achieved remarkable advancements. However, most of the neural MRC models are still far from robust and fail to generalize well in real-world applications. In order to comprehensively verify the robustness and generalization of MRC models, we introduce a real-world Chinese dataset -- DuReader_robust. It is designed to evaluate the MRC models from three aspects: over-sensitivity, over-stability and generalization. Comparing to previous work, the instances in DuReader_robust are natural texts, rather than the altered unnatural texts. It presents the challenges when applying MRC models to real-world applications. The experimental results show that MRC models do not perform well on the challenge test set. Moreover, we analyze the behavior of existing models on the challenge test set, which may provide suggestions for future model development. The dataset and codes are publicly available at https://github.com/baidu/DuReader.

cs.CL↗

arXiv2021

OPT: Omni-Perception Pre-Trainer for Cross-Modal Understanding and Generation

In this paper, we propose an Omni-perception Pre-Trainer (OPT) for cross-modal understanding and generation, by jointly modeling visual, text and audio resources. OPT is constructed in an encoder-decoder framework, including three single-modal encoders to generate token-based embeddings for each modality, a cross-modal encoder to encode the correlations among the three modalities, and two cross-modal decoders to generate text and image respectively. For the OPT's pre-training, we design a multi-task pretext learning scheme to model multi-modal resources from three different data granularities, \ie, token-, modality-, and sample-level modeling, through which OPT learns to align and translate among different modalities. The pre-training task is carried out on a large amount of image-text-audio triplets from Open Images. Experimental results show that OPT can learn strong image-text-audio multi-modal representations and achieve promising results on a variety of cross-modal understanding and generation tasks.

cs.CV↗

arXiv2021

Quantum metrology with precision reaching beyond-$1/N$ scaling through $N$-probe entanglement generating interactions

Nonlinear interactions are recognized as potential resources for quantum metrology, facilitating parameter estimation precisions that scale as the exponential Heisenberg limit of $2^{-N}$. We explore such nonlinearity and propose an associated quantum measurement scenario based on the nonlinear interaction of $N$-probe entanglement generating form. This scenario provides an enhanced precision scaling of $D^{-N}/(N-1)!$ with $D > 2$ a tunable parameter. In addition, it can be readily implemented in a variety of experimental platforms and applied to measurements of a wide range of quantities, including local gravitational acceleration $g$, magnetic field, and its higher-order gradients.

quant-ph↗

arXiv2021

Exploiting Large-scale Teacher-Student Training for On-device Acoustic Models

We present results from Alexa speech teams on semi-supervised learning (SSL) of acoustic models (AM) with experiments spanning over 3000 hours of GPU time, making our study one of the largest of its kind. We discuss SSL for AMs in a small footprint setting, showing that a smaller capacity model trained with 1 million hours of unsupervised data can outperform a baseline supervised system by 14.3% word error rate reduction (WERR). When increasing the supervised data to seven-fold, our gains diminish to 7.1% WERR; to improve SSL efficiency at larger supervised data regimes, we employ a step-wise distillation into a smaller model, obtaining a WERR of 14.4%. We then switch to SSL using larger student models in low data regimes; while learning efficiency with unsupervised data is higher, student models may outperform teacher models in such a setting. We develop a theoretical sketch to explain this behavior.

cs.SD↗

arXiv2021

Measuring Conversational Uptake: A Case Study on Student-Teacher Interactions

In conversation, uptake happens when a speaker builds on the contribution of their interlocutor by, for example, acknowledging, repeating or reformulating what they have said. In education, teachers' uptake of student contributions has been linked to higher student achievement. Yet measuring and improving teachers' uptake at scale is challenging, as existing methods require expensive annotation by experts. We propose a framework for computationally measuring uptake, by (1) releasing a dataset of student-teacher exchanges extracted from US math classroom transcripts annotated for uptake by experts; (2) formalizing uptake as pointwise Jensen-Shannon Divergence (pJSD), estimated via next utterance classification; (3) conducting a linguistically-motivated comparison of different unsupervised measures and (4) correlating these measures with educational outcomes. We find that although repetition captures a significant part of uptake, pJSD outperforms repetition-based baselines, as it is capable of identifying a wider range of uptake phenomena like question answering and reformulation. We apply our uptake measure to three different educational datasets with outcome indicators. Unlike baseline measures, pJSD correlates significantly with instruction quality in all three, providing evidence for its generalizability and for its potential to serve as an automated professional development tool for teachers.

cs.CL↗