arXiv subjects

Tao Wang

Publications and source records attributed to Tao Wang.

At least 559 records · Page 31Linked to original sources

arXiv2022

Density Shift of Optical Lattice Clock via Multi-Bands Sampling Exact Diagonalization Method

Density shift plays one of the major roles in the uncertainty of optical lattice clock, thus has attracted lots of theoretical and experimental studies. However, most of the theoretical research considered the single-band and collective approximation, so the density shift of the system at higher temperatures can not be analyzed accurately. Here, we design a numerical algorithm that combines Monte Carlo sampling and exact diagonalization and name it as Multi-Band Sampling Exact Diagonalization (MBSED). The MBSED method considers the collision of atoms between multi-bands, so it can provide the density shift of an optical lattice clock with higher precision. Such an algorithm will benefit the numerical simulation of an optical lattice clock, and may also be used in other platforms of quantum metrology. In addition, through our numerical simulation, we also found that the density shift of Rabi spectrum is slightly non-linear with atom number.

quant-ph↗

arXiv2022

The UV Excesses of Supernovae and the Implications for Studying Supernovae and Other Optical Transients

Supernovae (SNe), kilonovae (KNe), tidal disruption events (TDEs), optical afterglows of gamma ray bursts (GRBs), and many other optical transients are important phenomena in time-domain astronomy. Fitting the multi-band light curves (LCs) or the synthesized (pseudo-)bolometric LCs can be used to constrain the physical properties of optical transients. The (UV absorbed) blackbody module is one of the most important modules used to fit the multi-band LCs of optical transients having (UV absorbed) blackbody spectral energy distributions (SEDs). We find, however, that the SEDs of some SNe show UV excesses, which cannot be fitted by the model including a (UV absorbed) blackbody module. We construct the bolometric LCs and employ the (cooling plus) \Ni model to fit the constructed bolometric LCs, obtaining decent fits. Our results demonstrate that the optical transients showing UV excesses cannot be fitted by the multi-band models that include (UV-absorbed) blackbody module, but can be well modeled by constructing and fitting their bolometric LCs.

astro-ph.HE↗

arXiv2022

Boosting R-CNN: Reweighting R-CNN Samples by RPN's Error for Underwater Object Detection

Complicated underwater environments bring new challenges to object detection, such as unbalanced light conditions, low contrast, occlusion, and mimicry of aquatic organisms. Under these circumstances, the objects captured by the underwater camera will become vague, and the generic detectors often fail on these vague objects. This work aims to solve the problem from two perspectives: uncertainty modeling and hard example mining. We propose a two-stage underwater detector named boosting R-CNN, which comprises three key components. First, a new region proposal network named RetinaRPN is proposed, which provides high-quality proposals and considers objectness and IoU prediction for uncertainty to model the object prior probability. Second, the probabilistic inference pipeline is introduced to combine the first-stage prior uncertainty and the second-stage classification score to model the final detection score. Finally, we propose a new hard example mining method named boosting reweighting. Specifically, when the region proposal network miscalculates the object prior probability for a sample, boosting reweighting will increase the classification loss of the sample in the R-CNN head during training, while reducing the loss of easy samples with accurately estimated priors. Thus, a robust detection head in the second stage can be obtained. During the inference stage, the R-CNN has the capability to rectify the error of the first stage to improve the performance. Comprehensive experiments on two underwater datasets and two generic object detection datasets demonstrate the effectiveness and robustness of our method.

cs.CV↗

arXiv2022

Restricted phase space thermodynamics for black holes in higher dimensions and higher curvature gravities

The recently proposed restricted phase space thermodynamics is shown to be applicable to a large class of higher dimensional higher curvature gravity models coupled to Maxwell field, which are known as black hole scan models and are labeled by the spacetime dimension $d$ and the highest order $k$ of the Lanczos-Lovelock densities appearing in the action. Three typical example cases with $(d,k)=(5,1), (5,2)$ and $(6,2)$ are chosen as example cases and studied in some detail. These cases are representatives of Einstein-Hilbert, Chern-Simons and Born-Infield like gravity models. Our study indicates that the Einstein-Hilbert and Born-Infield like gravity models have similar thermodynamic behaviors, e.g. the existence of isocharge $T-S$ phase transitions with the same critical exponents, the existence of isovoltage $T-S$ transitions and the Hawking-Page like transitions, and the similar high temperature asymptotic behaviors for the isocharge heat capacities, etc. However, the Chern-Simons like $(5,2)$-model behaves quite differently. Neither isocharge nor isovoltage $T-S$ transitions could occur and no Hawking-Page like transition is allowed. This seems to indicate that the Einstein-Hilbert and Born-Infield like models belong to the same universality class while the Chern-Simons like models do not.

hep-th↗

arXiv2022

DLUNet: Semi-supervised Learning based Dual-Light UNet for Multi-organ Segmentation

The manual ground truth of abdominal multi-organ is labor-intensive. In order to make full use of CT data, we developed a semi-supervised learning based dual-light UNet. In the training phase, it consists of two light UNets, which make full use of label and unlabeled data simultaneously by using consistent-based learning. Moreover, separable convolution and residual concatenation was introduced light UNet to reduce the computational cost. Further, a robust segmentation loss was applied to improve the performance. In the inference phase, only a light UNet is used, which required low time cost and less GPU memory utilization. The average DSC of this method in the validation set is 0.8718. The code is available in https://github.com/laihaoran/Semi-SupervisednnUNet.

eess.IV↗

arXiv2022

ITA: Image-Text Alignments for Multi-Modal Named Entity Recognition

Recently, Multi-modal Named Entity Recognition (MNER) has attracted a lot of attention. Most of the work utilizes image information through region-level visual representations obtained from a pretrained object detector and relies on an attention mechanism to model the interactions between image and text representations. However, it is difficult to model such interactions as image and text representations are trained separately on the data of their respective modality and are not aligned in the same space. As text representations take the most important role in MNER, in this paper, we propose {\bf I}mage-{\bf t}ext {\bf A}lignments (ITA) to align image features into the textual space, so that the attention mechanism in transformer-based pretrained textual embeddings can be better utilized. ITA first aligns the image into regional object tags, image-level captions and optical characters as visual contexts, concatenates them with the input texts as a new cross-modal input, and then feeds it into a pretrained textual embedding model. This makes it easier for the attention module of a pretrained textual embedding model to model the interaction between the two modalities since they are both represented in the textual space. ITA further aligns the output distributions predicted from the cross-modal input and textual input views so that the MNER model can be more practical in dealing with text-only inputs and robust to noises from images. In our experiments, we show that ITA models can achieve state-of-the-art accuracy on multi-modal Named Entity Recognition datasets, even without image information.

cs.CL↗

arXiv2022

A No-reference Quality Assessment Metric for Point Cloud Based on Captured Video Sequences

Point cloud is one of the most widely used digital formats of 3D models, the visual quality of which is quite sensitive to distortions such as downsampling, noise, and compression. To tackle the challenge of point cloud quality assessment (PCQA) in scenarios where reference is not available, we propose a no-reference quality assessment metric for colored point cloud based on captured video sequences. Specifically, three video sequences are obtained by rotating the camera around the point cloud through three specific orbits. The video sequences not only contain the static views but also include the multi-frame temporal information, which greatly helps understand the human perception of the point clouds. Then we modify the ResNet3D as the feature extraction model to learn the correlation between the capture videos and corresponding subjective quality scores. The experimental results show that our method outperforms most of the state-of-the-art full-reference and no-reference PCQA metrics, which validates the effectiveness of the proposed method.

eess.IV↗

arXiv2022

Sub-percentage measure of distances to redshift of 0.1 by a new cosmic ruler

Distance-redshift diagrams probe expansion history of the Universe. We show that the stellar mass-binding energy (massE) relation of galaxies proposed in our previous study offers a new distance ruler at cosmic scales. By using elliptical galaxies in the main galaxy sample of the Sloan Digital Sky Survey Data Release 7, we construct a distance-redshift diagram over the redshift range from 0.05 to 0.2 with the massE ruler. The best-fit dark energy density is 0.675+-0.079 for flat Lambda-CDM, consistent with those by other probes. At the median redshift of 0.11, the median distance is estimated to have a fractional error of 0.34%, much lower than those by supernova (SN) Ia and baryonic acoustic oscillation (BAO) and even exceeding their future capability at this redshift. The above low-z measurement is useful for probing dark energy that dominates at the late Universe. For a flat dark energy equation of state model (flat wCDM), the massE alone constrains w to an error that is only a factor of 2.2, 1.7 and 1.3 times larger than those by BAO, SN Ia, and cosmic microwave background (CMB), respectively.

astro-ph.CO↗

arXiv2022

An Initial Investigation for Detecting Vocoder Fingerprints of Fake Audio

Many effective attempts have been made for fake audio detection. However, they can only provide detection results but no countermeasures to curb this harm. For many related practical applications, what model or algorithm generated the fake audio also is needed. Therefore, We propose a new problem for detecting vocoder fingerprints of fake audio. Experiments are conducted on the datasets synthesized by eight state-of-the-art vocoders. We have preliminarily explored the features and model architectures. The t-SNE visualization shows that different vocoders generate distinct vocoder fingerprints.

cs.SD↗

arXiv2022

Uniqueness, symmetry and convergence of positive ground state solutions of the Choquard type equation on a ball

This paper is concerned with the qualitative properties of the positive ground state solutions to the nonlocal Choquard type equation on a ball $B_R$. First, we prove the radial symmetry of the positive ground state solutions by using Talenti's inequality. Next we develop Newton's Theorem and then resort to the contraction mapping principle to establish the uniqueness of the positive ground state solutions. Finally, by constructing cut-off functions and applying energy comparison method, we show the convergence of the positive ground state solutions as $R\to \infty$. Our results generalize and improve the existing ones in the literature.

math.AP↗

arXiv2022

Multiple Instance Neural Networks Based on Sparse Attention for Cancer Detection using T-cell Receptor Sequences

Early detection of cancers has been much explored due to its paramount importance in biomedical fields. Among different types of data used to answer this biological question, studies based on T cell receptors (TCRs) are under recent spotlight due to the growing appreciation of the roles of the host immunity system in tumor biology. However, the one-to-many correspondence between a patient and multiple TCR sequences hinders researchers from simply adopting classical statistical/machine learning methods. There were recent attempts to model this type of data in the context of multiple instance learning (MIL). Despite the novel application of MIL to cancer detection using TCR sequences and the demonstrated adequate performance in several tumor types, there is still room for improvement, especially for certain cancer types. Furthermore, explainable neural network models are not fully investigated for this application. In this article, we propose multiple instance neural networks based on sparse attention (MINN-SA) to enhance the performance in cancer detection and explainability. The sparse attention structure drops out uninformative instances in each bag, achieving both interpretability and better predictive performance in combination with the skip connection. Our experiments show that MINN-SA yields the highest area under the ROC curve (AUC) scores on average measured across 10 different types of cancers, compared to existing MIL approaches. Moreover, we observe from the estimated attentions that MINN-SA can identify the TCRs that are specific for tumor antigens in the same T cell repertoire.

stat.ML↗

arXiv2022

GLaM: Efficient Scaling of Language Models with Mixture-of-Experts

Scaling language models with more data, compute and parameters has driven significant progress in natural language processing. For example, thanks to scaling, GPT-3 was able to achieve strong results on in-context learning tasks. However, training these large dense models requires significant amounts of computing resources. In this paper, we propose and develop a family of language models named GLaM (Generalist Language Model), which uses a sparsely activated mixture-of-experts architecture to scale the model capacity while also incurring substantially less training cost compared to dense variants. The largest GLaM has 1.2 trillion parameters, which is approximately 7x larger than GPT-3. It consumes only 1/3 of the energy used to train GPT-3 and requires half of the computation flops for inference, while still achieving better overall zero-shot and one-shot performance across 29 NLP tasks.

cs.CL↗

arXiv2022

Dense Gas and Star Formation in Nearby Infrared Bright Galaxies: APEX survey of HCN and HCO+ J=2-1

Both Galactic and extragalactic studies on star formation suggest that stars form directly from dense molecular gas. To trace such high volume density gas, HCN and HCO+ J=1-0 have been widely used for their high dipole moments, relatively high abundances, and often being the strongest lines after CO. However, HCN and HCO+ J=1-0 emission could be arguably dominated by the gas components at low volume densities. HCN J=2-1 and HCO+ J=2-1, with more suitable critical densities and excitation requirements, would trace typical dense gas closely related to star formation. Here we report new observations of HCN J=2-1 and HCO+ J=2-1 towards 17 nearby infrared-bright galaxies with the APEX 12-m telescope. The correlation slopes between luminosities of HCN J=2-1, and HCO+ J=2-1 and total infrared emission are 1.03 +- 0.05 and 1.00 +- 0.05, respectively. The correlations of their surface densities, normalised with the area of radio/sub-millimeter continuum, show even tighter relations (Slopes: 0.99 +- 0.03 and 1.02 +- 0.03). The eight AGN-dominated galaxies show no significant difference from the eleven star-formation dominated galaxies in above relations. The average HCN/HCO+ ratios are 1.15 +- 0.26 and 0.98 +- 0.42 for AGN-dominated and star-formation dominated galaxies, respectively, without obvious dependencies on infrared luminosity, dust temperature, or infrared pumping. The Magellanic Clouds roughly follow the same correlations, expanding to eight orders of magnitude. On the other hand, ultra-luminous infrared galaxies with active galactic nucleus (AGN) systematically lay above the correlations, indicating potential biases introduced by AGNs.

astro-ph.GA↗

arXiv2022

On Mitigating Hard Clusters for Face Clustering

Face clustering is a promising way to scale up face recognition systems using large-scale unlabeled face images. It remains challenging to identify small or sparse face image clusters that we call hard clusters, which is caused by the heterogeneity, \ie, high variations in size and sparsity, of the clusters. Consequently, the conventional way of using a uniform threshold (to identify clusters) often leads to a terrible misclassification for the samples that should belong to hard clusters. We tackle this problem by leveraging the neighborhood information of samples and inferring the cluster memberships (of samples) in a probabilistic way. We introduce two novel modules, Neighborhood-Diffusion-based Density (NDDe) and Transition-Probability-based Distance (TPDi), based on which we can simply apply the standard Density Peak Clustering algorithm with a uniform threshold. Our experiments on multiple benchmarks show that each module contributes to the final performance of our method, and by incorporating them into other advanced face clustering methods, these two modules can boost the performance of these methods to a new state-of-the-art. Code is available at: https://github.com/echoanran/On-Mitigating-Hard-Clusters.

cs.CV↗

arXiv2022

Variable degeneracy of graphs with restricted structures

Bernshteyn and Lee defined a new notion, weak degeneracy, which is slightly weaker than the ordinary degeneracy. It is proved that strictly $f$-degenerate transversal is a common generalization of list coloring, $L$-forested-coloring and DP-coloring. In this paper, we consider three classes of graphs, including planar graphs without any configuration in Fig. 2, toroidal graphs without any configuration in Fig. 5, and planar graphs without intersecting $5$-cycles. We give structural results for each class of graphs, and prove each structure is reducible for weakly $3$-degenerate and the existence of strictly $f$-degenerate transversals. As consequences, these three classes of graphs are weakly $3$-degenerate, and have a strictly $f$-degenerate transversal. Then these three classes of graph have DP-paint number at most four, and have list vertex arboricity at most two. This greatly improve all the results in [2-4, 11-13, 16-18, 22, 25, 32, 34]. Furthermore, the first and the third classes of graphs have Alon-Tarsi number at most four.

math.CO↗

arXiv2022

Low star-formation activity and low gas content of quiescent galaxies at $z=$ 3.5-4.0 constrained with ALMA

The discovery in deep near-infrared surveys of a population of massive quiescent galaxies at $z>3$ has given rise to the question of how they came to be quenched so early in the history of the Universe. Measuring their molecular gas properties can distinguish between physical processes where they stop forming stars due to a lack of fuel versus those where star-formation efficiency is reduced and the gas is retained. We conducted Atacama Large Millimeter/sub-millimeter Array (ALMA) observations of four quiescent galaxies at $z=$ 3.5-4.0 found by the Fourstar Galaxy Evolution Survey (ZFOURGE) and a serendipitous optically dark galaxy at $z=3.71$. We aim to investigate the presence of dust-obscured star-formation and their gas content by observing the dust continuum emission at Band-7 and the atomic carbon [C I]($^3P_1$-$^3P_0$) line at 492.16 GHz. Among the four quiescent galaxies, only one source is detected in the dust continuum at $λ_{\rm obs} = 870 {\rm μm}$. The sub-mm observations confirm their passive nature, and all of them are located more than four times below the main sequence of star-forming galaxies at $z=3.7$. None of the targets are detected in [C I], constraining their gas mass fractions to be $<$ 20%. These gas mass fractions are more than three times lower than the scaling relation for star-forming galaxies at $z=3.7$. These results support scenarios where massive galaxies at $z=$ 3.5-4.0 quench by consuming/expelling all the gas rather than by reducing the efficiency of the conversion of their gas into stars.

astro-ph.GA↗

arXiv2022

DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition

The MultiCoNER shared task aims at detecting semantically ambiguous and complex named entities in short and low-context settings for multiple languages. The lack of contexts makes the recognition of ambiguous named entities challenging. To alleviate this issue, our team DAMO-NLP proposes a knowledge-based system, where we build a multilingual knowledge base based on Wikipedia to provide related context information to the named entity recognition (NER) model. Given an input sentence, our system effectively retrieves related contexts from the knowledge base. The original input sentences are then augmented with such context information, allowing significantly better contextualized token representations to be captured. Our system wins 10 out of 13 tracks in the MultiCoNER shared task.

cs.CL↗

arXiv2022

Theoretical and experimental study on Noise Equivalent Power of X-ray semiconductor ultra-fast response material based on the rad-optic effect

Semiconductor material based on the rad-optic effect enables ultra-fast detection of X-rays and plays an important role in fusion diagnostics. Obtaining the accurate noise equivalent power (NEP) of the semiconductor ultrafast response material is the key to detecting X-rays. In this paper, the refractive index change mechanism of the semiconductor under X-ray irradiation was analyzed, and the quantitative relationship between the diffraction efficiency and the X-ray photon energy was established through the LT-AlGaAs diffraction imaging experiments. The impulse responses of LT-AlGaAs under 1 KeV-10 KeV X-ray radiation were calculated, revealing the variation of NEP density with radiated photon energy. In the case of bombarding the Al target to generate 1.5 KeV X-rays, the imaging experiments of LT-AlGaAs were performed. The diffraction image of LT-AlGaAs has a linear relationship with the radiation intensity, and the NEP density of LT-AlGaAs reaches 4.80*105W/cm2. This study has reference significance for the development of ultra-fast X-ray imaging systems based on the rad-optic effect.

physics.ins-det↗