arXiv subjects

Zhuo Li

Publications and source records attributed to Zhuo Li.

At least 145 records · Page 8Linked to original sources

arXiv2023

Implant Global and Local Hierarchy Information to Sequence based Code Representation Models

Source code representation with deep learning techniques is an important research field. There have been many studies that learn sequential or structural information for code representation. But sequence-based models and non-sequence-models both have their limitations. Researchers attempt to incorporate structural information to sequence-based models, but they only mine part of token-level hierarchical structure information. In this paper, we analyze how the complete hierarchical structure influences the tokens in code sequences and abstract this influence as a property of code tokens called hierarchical embedding. The hierarchical embedding is further divided into statement-level global hierarchy and token-level local hierarchy. Furthermore, we propose the Hierarchy Transformer (HiT), a simple but effective sequence model to incorporate the complete hierarchical embeddings of source code into a Transformer model. We demonstrate the effectiveness of hierarchical embedding on learning code structure with an experiment on variable scope detection task. Further evaluation shows that HiT outperforms SOTA baseline models and show stable training efficiency on three source code-related tasks involving classification and generation tasks across 8 different datasets.

cs.SE↗

arXiv2023

PCF: ECAPA-TDNN with Progressive Channel Fusion for Speaker Verification

ECAPA-TDNN is currently the most popular TDNN-series model for speaker verification, which refreshed the state-of-the-art(SOTA) performance of TDNN models. However, one-dimensional convolution has a global receptive field over the feature channel. It destroys the time-frequency relevance of the spectrogram. Besides, as ECAPA-TDNN only has five layers, a much shallower structure compared to ResNet restricts the capability to generate deep representations. To further improve ECAPA-TDNN, we propose a progressive channel fusion strategy that splits the spectrogram across the feature channel and gradually expands the receptive field through the network. Secondly, we enlarge the model by extending the depth and adding branches. Our proposed model achieves EER with 0.718 and minDCF(0.01) with 0.0858 on vox1o, relatively improved 16.1\% and 19.5\% compared with ECAPA-TDNN-large.

cs.SD↗

arXiv2023

Neural Episodic Control with State Abstraction

Existing Deep Reinforcement Learning (DRL) algorithms suffer from sample inefficiency. Generally, episodic control-based approaches are solutions that leverage highly-rewarded past experiences to improve sample efficiency of DRL algorithms. However, previous episodic control-based approaches fail to utilize the latent information from the historical behaviors (e.g., state transitions, topological similarities, etc.) and lack scalability during DRL training. This work introduces Neural Episodic Control with State Abstraction (NECSA), a simple but effective state abstraction-based episodic control containing a more comprehensive episodic memory, a novel state evaluation, and a multi-step state analysis. We evaluate our approach to the MuJoCo and Atari tasks in OpenAI gym domains. The experimental results indicate that NECSA achieves higher sample efficiency than the state-of-the-art episodic control-based approaches. Our data and code are available at the project website\footnote{\url{https://sites.google.com/view/drl-necsa}}.

cs.LG↗

arXiv2023

Numerical study of synchrotron and inverse-Compton radiation from gamma-ray burst afterglows with decaying microturbulence

The multiwavelength observations of GRB afterglows, together with some high-performance particle-in-cell simulations, hint that the magnetic field may decay behind the shock front. In this work, we develop a numerical code to calculate the evolution of the accelerated electron distribution, their synchrotron and inverse-Compton (IC) spectra and accordingly the light curves (LCs) under the assumption of decaying microturbulence (DM) downstream of the shock, $ε_B(t_p')\propto t_p'^{α_t}$ with $t_p'$ the fluid proper time since injection. We find: (1) The synchrotron spectrum in the DM model is similar to that in the homogeneous turbulence (HT) model with very low magnetic field strength. However, the difference in the IC spectral component is relatively more obvious between them, due to the significant change of the postshock electron energy distribution with DM. (2) If the magnetic field decay faster, there are less electrons cool fast, and the IC spectral component becomes weaker. (3) The LCs in the DM model decay steeper than in the HT model, and the spectral evolution and the LCs in the DM model is similar to the HT model where the magnetic field energy fraction decreases with observer time, $ε_B(t) \propto t^{5α_t /8}$. (4) The DM model can naturally produce a significant IC spectral component in TeV energy range, but due to the Klein-Nishina suppression the IC power cannot be far larger than the synchrotron power. We apply the DM model to describe the afterglow data of GRB 190114C and find the magnetic field decay exponent $α_t\sim -0.4$ and the electron spectral index $p\sim2.4$. Future TeV observations of the IC emission from GRB afterglows will further help to probe the poorly known microphysics of relativistic shocks.

astro-ph.HE↗

arXiv2023

EdgeYOLO: An Edge-Real-Time Object Detector

This paper proposes an efficient, low-complexity and anchor-free object detector based on the state-of-the-art YOLO framework, which can be implemented in real time on edge computing platforms. We develop an enhanced data augmentation method to effectively suppress overfitting during training, and design a hybrid random loss function to improve the detection accuracy of small objects. Inspired by FCOS, a lighter and more efficient decoupled head is proposed, and its inference speed can be improved with little loss of precision. Our baseline model can reach the accuracy of 50.6% AP50:95 and 69.8% AP50 in MS COCO2017 dataset, 26.4% AP50:95 and 44.8% AP50 in VisDrone2019-DET dataset, and it meets real-time requirements (FPS>=30) on edge-computing device Nvidia Jetson AGX Xavier. We also designed lighter models with less parameters for edge computing devices with lower computing power, which also show better performances. Our source code, hyper-parameters and model weights are all available at https://github.com/LSH9832/edgeyolo.

cs.CV↗

arXiv2023

Time-attenuating Twin Delayed DDPG Reinforcement Learning for Trajectory Tracking Control of Quadrotors

Continuous trajectory tracking control of quadrotors is complicated when considering noise from the environment. Due to the difficulty in modeling the environmental dynamics, tracking methodologies based on conventional control theory, such as model predictive control, have limitations on tracking accuracy and response time. We propose a Time-attenuating Twin Delayed DDPG, a model-free algorithm that is robust to noise, to better handle the trajectory tracking task. A deep reinforcement learning framework is constructed, where a time decay strategy is designed to avoid trapping into local optima. The experimental results show that the tracking error is significantly small, and the operation time is one-tenth of that of a traditional algorithm. The OpenAI Mujoco tool is used to verify the proposed algorithm, and the simulation results show that, the proposed method can significantly improve the training efficiency and effectively improve the accuracy and convergence stability.

cs.RO↗

arXiv2023

Simulation study on the optical processes at deep-sea neutrino telescope sites

The performance of a large-scale water Cherenkov neutrino telescope relies heavily on the transparency of the surrounding water, quantified by its level of light absorption and scattering. A pathfinder experiment was carried out to measure the optical properties of deep seawater in South China Sea with light-emitting diodes (LEDs) as light sources, photon multiplier tubes (PMTs) and cameras as photon sensors. Here, we present an optical simulation program employing the Geant4 toolkit to understand the absorption and scattering processes in the deep seawater, which helps to extract the underlying optical properties from the experimental data. The simulation results are compared with the experimental data and show good agreements. We also verify the analysis methods that utilize various observables of the PMTs and the cameras with this simulation program, which can be easily adapted by other neutrino telescope pathfinder experiments and future large-scale detectors.

astro-ph.IM↗

arXiv2022

Jet-cloud/star interaction as an interpretation of neutrino outburst from the blazar TXS 0506+056

A neutrino outburst between September 2014 and March 2015 was discovered from the blazar TXS 0506+056 by an investigation of $9.5$ years of IceCube data, while the blazar was in a quiescent state during the outburst with a gamma-ray flux only about one-fifth of the neutrino flux. In this work, we give a possible interpretation of the abnormal feature by proposing that the neutrino outburst originates from the interaction between a relativistic jet and a dense gas cloud formed via the tidally disrupted envelope of a red giant being blown apart by the impact of the jet. Gamma-ray photons and electron/positron pairs that are produced through the hadronuclear interactions correspondingly will induce electromagnetic cascades and then make the cloud ionized and thermalized. The EM radiation from jet-cloud/star interaction is mainly contributed by the relatively low-energy relativistic protons which propagate in the diffusion regime inside the cloud due to magnetic deflections, whereas the observed high-energy neutrinos ($\gtrsim 100\,\rm TeV$) are produced by the relatively high-energy protons which can keep beamed owing to the weak magnetic deflections, inducing a much higher flux of neutrinos than electromagnetic radiation. The observed low-energy electromagnetic radiations during the neutrino outburst period are almost as same as that in the quiescent state of the source, so it may mainly arise as same as the general quiescent state. As a result, due to the intrusion of a dense cloud, the neutrino outburst can be expected, and in the meantime, the accompanying electromagnetic radiations from hadronic processes will not cause any enhancement in the blazar's electromagnetic flux.

astro-ph.HE↗

arXiv2022

Poison Attack and Defense on Deep Source Code Processing Models

In the software engineering community, deep learning (DL) has recently been applied to many source code processing tasks. Due to the poor interpretability of DL models, their security vulnerabilities require scrutiny. Recently, researchers have identified an emergent security threat, namely poison attack. The attackers aim to inject insidious backdoors into models by poisoning the training data with poison samples. Poisoned models work normally with clean inputs but produce targeted erroneous results with poisoned inputs embedded with triggers. By activating backdoors, attackers can manipulate the poisoned models in security-related scenarios. To verify the vulnerability of existing deep source code processing models to the poison attack, we present a poison attack framework for source code named CodePoisoner as a strong imaginary enemy. CodePoisoner can produce compilable even human-imperceptible poison samples and attack models by poisoning the training data with poison samples. To defend against the poison attack, we further propose an effective defense approach named CodeDetector to detect poison samples in the training data. CodeDetector can be applied to many model architectures and effectively defend against multiple poison attack approaches. We apply our CodePoisoner and CodeDetector to three tasks, including defect detection, clone detection, and code repair. The results show that (1) CodePoisoner achieves a high attack success rate (max: 100%) in misleading models to targeted erroneous behaviors. It validates that existing deep source code processing models have a strong vulnerability to the poison attack. (2) CodeDetector effectively defends against multiple poison attack approaches by detecting (max: 100%) poison samples in the training data. We hope this work can help practitioners notice the poison attack and inspire the design of more advanced defense techniques.

cs.SE↗

arXiv2022

Deepfake Detection System for the ADD Challenge Track 3.2 Based on Score Fusion

This paper describes the deepfake audio detection system submitted to the Audio Deep Synthesis Detection (ADD) Challenge Track 3.2 and gives an analysis of score fusion. The proposed system is a score-level fusion of several light convolutional neural network (LCNN) based models. Various front-ends are used as input features, including low-frequency short-time Fourier transform and Constant Q transform. Due to the complex noise and rich synthesis algorithms, it is difficult to obtain the desired performance using the training set directly. Online data augmentation methods effectively improve the robustness of fake audio detection systems. In particular, the reasons for the poor improvement of score fusion are explored through visualization of the score distributions and comparison with score distribution on another dataset. The overfitting of the model to the training set leads to extreme values of the scores and low correlation of the score distributions, which makes score fusion difficult. Fusion with partially fake audio detection system improves system performance further. The submission on track 3.2 obtained the weighted equal error rate (WEER) of 11.04\%, which is one of the best performing systems in the challenge.

eess.AS↗

arXiv2022

Active control of thermal emission by graphene-nanowire coupled plasmonic metasurfaces

Metasurfaces, together with graphene plasmonics, have become prominent for the emissivity control in thermal engineering, both passively through changing the geometric parameters and packing density of the metasurfaces, and actively through graphene gating or doping. We demonstrate a graphene-nanowire coupled plasmonic metasurface utilizing the hybrid localized surface plasmon modes of the nanowire array and graphene. The nanowire array makes the hybrid surface plasmon mode localized, allowing a free-space excitation. The single layer graphene, via the gating between the underneath mirror and a top electrode, can actively tune the spectral emissivity by almost 90%. In addition, the hybrid plasmon mode provides an extra degree of freedom to modulate the p-polarized emissivity with a five-fold enhancement, especially for large emission angles.

physics.app-ph↗

arXiv2022

Learning to Re-weight Examples with Optimal Transport for Imbalanced Classification

Imbalanced data pose challenges for deep learning based classification models. One of the most widely-used approaches for tackling imbalanced data is re-weighting, where training samples are associated with different weights in the loss function. Most of existing re-weighting approaches treat the example weights as the learnable parameter and optimize the weights on the meta set, entailing expensive bilevel optimization. In this paper, we propose a novel re-weighting method based on optimal transport (OT) from a distributional point of view. Specifically, we view the training set as an imbalanced distribution over its samples, which is transported by OT to a balanced distribution obtained from the meta set. The weights of the training samples are the probability mass of the imbalanced distribution and learned by minimizing the OT distance between the two distributions. Compared with existing methods, our proposed one disengages the dependence of the weight learning on the concerned classifier at each iteration. Experiments on image, text and point cloud datasets demonstrate that our proposed re-weighting method has excellent performance, achieving state-of-the-art results in many cases and providing a promising tool for addressing the imbalanced classification issue.

cs.LG↗

arXiv2022

Long-duration Gamma-ray Burst and Associated Kilonova Emission from Fast-spinning Black Hole--Neutron Star Mergers

Here we collect three unique bursts, GRBs\,060614, 211211A and 211227A, all characterized by a long-duration main emission (ME) phase and a rebrightening extended emission (EE) phase, to study their observed properties and the potential origin as neutron star-black hole (NSBH) mergers. NS-first-born (BH-first-born) NSBH mergers tend to contain fast-spinning (non-spinning) BHs that more easily (hardly) allow tidal disruption to happen with (without) forming electromagnetic signals. We find that NS-first-born NSBH mergers can well interpret the origins of these three GRBs, supported by that: (1) Their X-ray MEs and EEs show unambiguous fall-back accretion signatures, decreasing as $\propto{t}^{-5/3}$, which might account for their long duration. The EEs can result from the fall-back accretion of $r$-process heating materials, predicted to occur after NSBH mergers. (2) The beaming-corrected local event rate density for this type of merger-origin long-duration GRBs is $\mathcal{R}_0\sim2.4^{+2.3}_{-1.3}\,{\rm{Gpc}}^{-3}\,{\rm{yr}}^{-1}$, consistent with that of NS-first-born NSBH mergers. (3) Our detailed analysis on the EE, afterglow and kilonova of the recently high-impact event GRB\,211211A reveals it could be a merger between a $\sim1.23^{+0.06}_{-0.07}\,M_\odot$ NS and a $\sim8.21^{+0.77}_{-0.75}\,M_\odot$ BH with an aligned-spin of $χ_{\rm{BH}}\sim0.62^{+0.06}_{-0.07}$, supporting an NS-first-born NSBH formation channel. Long-duration burst with rebrightening fall-back accretion signature after ME, and bright kilonova might be commonly observed features for on-axis NSBHs. We estimate the multimessenger detection rate between gravitational waves, GRBs and kilonovae from NSBH mergers in O4 (O5) is $\sim0.1\,{\rm{yr}}^{-1}$ ($\sim1\,{\rm{yr}}^{-1}$).

astro-ph.HE↗

arXiv2022

The HCCL System for the NIST SRE21

This paper describes the systems developed by the HCCL team for the NIST 2021 speaker recognition evaluation (NIST SRE21).We first explore various state-of-the-art speaker embedding extractors combined with a novel circle loss to obtain discriminative deep speaker embeddings. Considering that cross-channel and cross-linguistic speaker recognition are the key challenges of SRE21, we introduce several techniques to reduce the cross-domain mismatch. Specifically, Codec and speech enhancement are directly applied to the raw speech to eliminate the codecs and the environment noise mismatch. We denote the methods that work directly on speech to eliminate the relatively explicit mismatches collectively as data adaptation methods. Experiments show that data adaption methods achieve 15\% improvements over our baseline. Furthermore, some popular back-ends domain adaptation algorithms are deployed on speaker embeddings to alleviate speaker performance degradation caused by the implicit mismatch. Score calibration is a major failure for us in SRE21. The reason is that score calibration with too many parameters easily lead to overfitting problems.

cs.SD↗

arXiv2022

SASV Based on Pre-trained ASV System and Integrated Scoring Module

Based on the assumption that there is a correlation between anti-spoofing and speaker verification, a Total-Divide-Total integrated Spoofing-Aware Speaker Verification (SASV) system based on pre-trained automatic speaker verification (ASV) system and integrated scoring module is proposed and submitted to the SASV 2022 Challenge. The training and scoring of ASV and anti-spoofing countermeasure (CM) in current SASV systems are relatively independent, ignoring the correlation. In this paper, by leveraging the correlation between the two tasks, an integrated SASV system can be obtained by simply training a few more layers on the basis of the baseline pre-trained ASV subsystem. The features in pre-trained ASV system are utilized for logical access spoofing speech detection. Further, speaker embeddings extracted by the pre-trained ASV system are used to improve the performance of the CM. The integrated scoring module takes the embeddings of the ASV and anti-spoofing branches as input and preserves the correlation between the two tasks through matrix operations to produce integrated SASV scores. Submitted primary system achieved equal error rate (EER) of 3.07\% on the development dataset of the SASV 2022 Challenge and 4.30\% on the evaluation part, which is a 25\% improvement over the baseline systems.

eess.AS↗

arXiv2022

Design and analysis of electrothermal metasurfaces

Electrothermal metasurfaces have attracted extensive attention due to their ability to dynamically control thermal infrared radiation. Although previous studies were mainly focused on the metasurfaces with infinite unit cells, the finite-size effect can be a critical design factor for developing thermal metasurfaces with fast response and broad temperature uniformity in practice. Here, we study the thermal metasurfaces consisting of gold nanorods with a finite array size, which, with only several periods, can achieve a resonance close to that of the infinite case. More importantly, such a small footprint due to the finite array size results in the response time down to a nanosecond level. Furthermore, the number of the unit cells in the direction perpendicular to the axis of the nanorods is found to be insensitive to the resonance and response time, thus providing a tunability in aspect ratio that can boost the temperature uniformity in the sub-Kelvin level.

physics.optics↗

arXiv2022

Could TDE outflows produce the PeV neutrino events?

A tidal disruption event (TDE), AT2019dsg, was observed to be associated with a PeV neutrino event, IceCube-191001A, lagging the optical outburst by a half year. It is known that TDEs may generate ultrafast outflows. If the TDE occurs in a cloudy environment, the outflow-cloud interactions may form shock waves which generate accelerated protons and hence delayed neutrinos from hadronic interactions in clouds. Here we investigate the neutrino production in AT2019dsg by examining the TDE outflow-cloud interaction model. We find that, for an outflow with a velocity of 0.07c and a kinetic luminosity of $10^{45}\rm erg\ s^{-1}$, protons may be accelerated up to $\sim$ 60 PeV by the bow shocks, and generate PeV neutrinos by interactions with clouds. The predicted neutrino number in this model depends on the uncertainties of model parameters and in order to match the observations, some challenging values of parameters have been involved. The PeV neutrino event number can be $\sim 4\times10^{-3}$ for a hard proton index $Γ=1.5$.

astro-ph.HE↗

arXiv2022

Graph Enhanced Contrastive Learning for Radiology Findings Summarization

The impression section of a radiology report summarizes the most prominent observation from the findings section and is the most important section for radiologists to communicate to physicians. Summarizing findings is time-consuming and can be prone to error for inexperienced radiologists, and thus automatic impression generation has attracted substantial attention. With the encoder-decoder framework, most previous studies explore incorporating extra knowledge (e.g., static pre-defined clinical ontologies or extra background information). Yet, they encode such knowledge by a separate encoder to treat it as an extra input to their models, which is limited in leveraging their relations with the original findings. To address the limitation, we propose a unified framework for exploiting both extra knowledge and the original findings in an integrated way so that the critical information (i.e., key words and their relations) can be extracted in an appropriate way to facilitate impression generation. In detail, for each input findings, it is encoded by a text encoder, and a graph is constructed through its entities and dependency tree. Then, a graph encoder (e.g., graph neural networks (GNNs)) is adopted to model relation information in the constructed graph. Finally, to emphasize the key words in the findings, contrastive learning is introduced to map positive samples (constructed by masking non-key words) closer and push apart negative ones (constructed by masking key words). The experimental results on OpenI and MIMIC-CXR confirm the effectiveness of our proposed method.

cs.CL↗