arXiv subjects

Shan Gao

Publications and source records attributed to Shan Gao.

At least 73 records · Page 4Linked to original sources

arXiv2022

Room geometry blind inference based on the localization of real sound source and first order reflections

The conventional room geometry blind inference techniques with acoustic signals are conducted based on the prior knowledge of the environment, such as the room impulse response (RIR) or the sound source position, which will limit its application under unknown scenarios. To solve this problem, we have proposed a room geometry reconstruction method in this paper by using the geometric relation between the direct signal and first-order reflections. In addition to the information of the compact microphone array itself, this method does not need any precognition of the environmental parameters. Besides, the learning-based DNN models are designed and used to improve the accuracy and integrity of the localization results of the direct source and first-order reflections. The direction of arrival (DOA) and time difference of arrival (TDOA) information of the direct and reflected signals are firstly estimated using the proposed DCNN and TD-CNN models, which have higher sensitivity and accuracy than the conventional methods. Then the position of the sound source is inferred by integrating the DOA, TDOA and array height using the proposed DNN model. After that, the positions of image sources and corresponding boundaries are derived based on the geometric relation. Experimental results of both simulations and real measurements verify the effectiveness and accuracy of the proposed techniques compared with the conventional methods under different reverberant environments.

cs.SD↗

arXiv2022

Left Ventricle Contouring of Apical Three-Chamber Views on 2D Echocardiography

We propose a new method to automatically contour the left ventricle on 2D echocardiographic images. Unlike most existing segmentation methods, which are based on predicting segmentation masks, we focus at predicting the endocardial contour and the key landmark points within this contour (basal points and apex). This provides a representation that is closer to how experts perform manual annotations and hence produce results that are physiologically more plausible. Our proposed method uses a two-headed network based on the U-Net architecture. One head predicts the 7 contour points, and the other head predicts a distance map to the contour. This approach was compared to the U-Net and to a point based approach, achieving performance gains of up to 30\% in terms of landmark localisation (<4.5mm) and distance to the ground truth contour (<3.5mm).

eess.IV↗

arXiv2022

Bone tumor suppression in rabbits by hyperthermia below the clinical safety limit using aligned magnetic bone cement

Demonstrating highly efficient alternating current (AC) magnetic field heating of nanoparticles in physiological environments under clinically safe field parameters has remained a great challenge, hindering clinical applications of magnetic hyperthermia. In this work, we report exceptionally high loss power of magnetic bone cement under clinical safety limit of AC field parameters, incorporating DC field-aligned soft magnetic Zn0.3Fe2.7O4 nanoparticles with low concentration. Under an AC field of 4 kA/m at 430 kHz, the aligned bone cement with 0.2 wt% nanoparticles achieved a temperature increase of 30 C in 180 s. This amounts to a specific loss power value of 327 W/gmetal and an intrinsic loss power of 47 nHm^2/kg, which is enhanced by 50-fold compared to randomly oriented samples. The high-performance magnetic bone cement allows for the demonstration of effective hyperthermia suppression of tumor growth in the bone marrow cavity of New Zealand White Rabbits subjecting to rapid cooling due to blood circulation, and significant enhancement of survival rate.

physics.med-ph↗

arXiv2021

Direct source and early reflections localization using deep deconvolution network under reverberant environment

This paper proposes a deconvolution-based network (DCNN) model for DOA estimation of direct source and early reflections under reverberant scenarios. Considering that the first-order reflections of the sound source also contain spatial directivity like the direct source, we treat both of them as the sources in the learning process. We use the covariance matrix of high order Ambisonics (HOA) signals in the time domain as the input feature of the network, which is concise while containing precise spatial information under reverberant scenarios. Besides, we use the deconvolution-based network for the spatial pseudo-spectrum (SPS) reconstruction in the 2D polar space, based on which the spatial relationship between elevation and azimuth can be depicted. We have carried out a series of experiments based on simulated and measured data under different reverberant scenarios, which prove the robustness and accuracy of the proposed DCNN model.

eess.AS↗

arXiv2021

Contrastive Learning for View Classification of Echocardiograms

Analysis of cardiac ultrasound images is commonly performed in routine clinical practice for quantification of cardiac function. Its increasing automation frequently employs deep learning networks that are trained to predict disease or detect image features. However, such models are extremely data-hungry and training requires labelling of many thousands of images by experienced clinicians. Here we propose the use of contrastive learning to mitigate the labelling bottleneck. We train view classification models for imbalanced cardiac ultrasound datasets and show improved performance for views/classes for which minimal labelled data is available. Compared to a naive baseline model, we achieve an improvement in F1 score of up to 26% in those views while maintaining state-of-the-art performance for the views with sufficiently many labelled training observations.

cs.CV↗

arXiv2021

The Blazar Hadronic Code Comparison Project

Blazar hadronic models have been developed in the past decades as an alternative to leptonic ones. In hadronic models the gamma-ray emission is associated with synchrotron emission by protons, and/or secondary leptons produced in proton-photon interactions. Together with photons, hadronic emission models predict the emission of neutrinos that are therefore the smoking gun for acceleration of relativistic hadrons in blazar jets. The simulation of proton-photon interactions and all associated radiative processes is a complex numerical task, and different approaches to the problem have been adopted in the literature. So far, no systematic comparison between the different codes has been performed, preventing a clear understanding of the underlying uncertainties in the numerical simulations. To fill this gap, we have undertaken the first comprehensive comparison of blazar hadronic codes, and the results from this effort will be presented in this contribution.

astro-ph.HE↗

arXiv2020

Multi-wavelength and neutrino emission from blazar PKS 1502+106

In July of 2019, the IceCube experiment detected a high-energy neutrino from the direction of the powerful blazar PKS 1502+106. We perform multi-wavelength and multi-messenger modeling of this source, using a fully self-consistent one-zone model that includes the contribution of external radiation fields typical of flat-spectrum radio quasars (FSRQs). We identify three different activity states of the blazar: one quiescent state and two flaring states with hard and soft gamma-ray spectra. We find two hadronic models that can describe the multi-wavelength emission during all three states: a leptohadronic model with a contribution from photo-hadronic processes to X-rays and gamma rays, and a proton synchrotron model, where the emission from keV to 10 GeV comes from proton synchrotron radiation. Both models predict a substantial neutrino flux that is correlated with the gamma-ray and soft X-ray fluxes. Our results are compatible with the detection of a neutrino during the quiescent state, based on event rate statistics. We conclude that the soft X-ray spectra observed during bright flares strongly suggest a hadronic contribution, which can be interpreted as additional evidence for cosmic ray acceleration in the source independently of neutrino observations. We find that more arguments can be made in favor of the leptohadronic model vis-a-vis the proton synchrotron scenario, such as a lower energetic demand during the quiescent state. However, the same leptohadronic model would be disfavored for flaring states of PKS 1502+106 if no IceCube events were found from the direction of the source before 2010, which would require an archival search.

astro-ph.HE↗

arXiv2020

A puzzle for the field ontologists

It has been widely thought that the wave function describes a real, physical field in a realist interpretation of quantum mechanics. In this paper, I present a new analysis of the field ontology for the wave function. First, I argue that the non-existence of self-interactions for a quantum system such as an electron poses a puzzle for the field ontologists. If the wave function represents a physical field, then it seems odd that there are (electromagnetic and gravitational) interactions between the fields of two electrons but no interactions between two parts of the field of an electron. Next, I argue that the three solutions a field ontologist may provide are not fully satisfactory. Finally, I propose a solution of this puzzle that leads to a particle ontological interpretation of the wave function.

quant-ph↗

arXiv2020

The measurement problem revisited

It has been realized that the measurement problem of quantum mechanics is essentially the determinate-experience problem, and in order to solve the problem, the physical state representing the measurement result is required to be also the physical state on which the mental state of an observer supervenes. This necessitates a systematic analysis of the forms of psychophysical connection in the solutions to the measurement problem. In this paper, I propose a new, mentalistic formulation of the measurement problem which lays more stress on psychophysical connection. By this new formulation, it can be seen more clearly that the three main solutions to the measurement problem, namely Everett's theory, Bohm's theory and collapse theories, correspond to three different forms of psychophysical connection. I then analyze these forms of psychophysical connection. It is argued that the forms of psychophysical connection required by Everett's and Bohm's theories have potential problems, while an analysis of how the mental state of an observer supervenes on her wave function may help solve the structured tails problem of collapse theories.

quant-ph↗

arXiv2020

Protective Measurements and the Reality of the Wave Function

It has been debated whether protective measurement implies the reality of the wave function. In this paper, I present a new analysis of the relationship between protective measurements and the reality of the wave function. First, I briefly introduce protective measurements and the ontological models framework for them. Second, I give a simple proof of Hardy's theorem in terms of protective measurements. Third, I analyze two suggested $ψ$-epistemic models of a protective measurement. It is shown that although these models can explain the appearance of expectation values of observables in a single measurement, their predictions about the variance of the result of a non-ideal protective measurement are different from those of quantum mechanics. Finally, I argue that under an auxiliary finiteness assumption about the dynamics of the ontic state, protective measurement implies the reality of the wave function in the ontological models framework.

quant-ph↗

arXiv2019

Multi-messenger interpretation of the neutrinos from TXS 0506+056

We discuss possible interpretations of the neutrinos observed from the AGN blazar TXS 0506+056 in the multi-messenger and multi-wavelength context, including both the 2014-15 and 2017 neutrino flares. While the neutrino observed in September 2017 has to describe contemporary data in e.g. the X-ray and VHE gamma-ray ranges, data at the 2014-15 excess are much sparser. We demonstrate that in both cases the simplest possible one-zone AGN blazar models face challenges. While the 2017 flare can be well interpreted by considering more sophisticated source geometries, the 2014-15 flare is much harder to describe with conventional models. One challenge is the energy injected into the electromagnetic cascade coming together with the neutrino production, which cannot be reconciled with the 13 observed neutrino events. We also speculate if a common interpretation of both flares is feasible.

astro-ph.HE↗

arXiv2019

Modelling the coincident observation of a high-energy neutrino and a bright blazar flare

In September 2017, the IceCube Neutrino Observatory recorded a very-high-energy neutrino in directional coincidence with a blazar in an unusually bright gamma-ray state, TXS0506+056. Blazars are prominent photon sources in the universe because they harbor a relativistic jet whose radiation is strongly collimated and amplified. High-energy atomic nuclei known as cosmic rays can produce neutrinos; thus the recent detection may help identifying the sources of the diffuse neutrino flux and the energetic cosmic rays. Here we report on a self-consistent analysis of the physical relation between the observed neutrino and the blazar, in particular the time evolution and spectral behavior of neutrino and photon emission. We demonstrate that a moderate enhancement in the number of cosmic rays during the flare can yield a very strong increase of the neutrino flux which is limited by co-produced hard X-rays and TeV gamma rays. We also test typical radiation models for compatibility and identify several model classes as incompatible with the observations. We investigate to what degree the findings can be generalized to the entire population of blazars, to determine the relation between their output in photons, neutrinos, and cosmic rays, and suggest how to optimize the strategy of future observations.

astro-ph.HE↗

arXiv2019

Leptohadronic blazar models applied to the 2014-15 flare of TXS 0506+056

We investigate whether the emission of neutrinos observed in 2014-15 from the direction of the blazar TXS 0506+056 can be accommodated with leptohadronic multi-wavelength models of the source commonly adopted for the 2017 flare. While multi-wavelength data during the neutrino flare are sparse, the large number of neutrino events ($13\pm5$) challenges the missing activity in gamma rays. We illustrate that two to five neutrino events during the flare can be explained with leptohadronic models of different categories: a one-zone model, a compact core model, and an external radiation field model. If, however, significantly more events were to be accommodated, the predicted multi-wavelength emission levels would be in conflict with observational X-ray constraints, or with the high-energy gamma ray fluxes observed by the Fermi LAT, depending on the model. For example, while the external radiation field model can predict up to five neutrino events without violating X-ray constraints, the absorption of high-energy gamma rays is in minor tension with data. We therefore do not find any model that can simultaneously explain the high event number quoted by IceCube and the (sparse) electromagnetic data during the neutrino flare.

astro-ph.HE↗

arXiv2019

Learning to Label Seismic Structures with Deconvolution Networks and Weak Labels

Recently, there has been increasing interest in using deep learning techniques for various seismic interpretation tasks. However, unlike shallow machine learning models, deep learning models are often far more complex and can have hundreds of millions of free parameters. This not only means that large amounts of computational resources are needed to train these models, but more critically, they require vast amounts of labeled training data as well. In this work, we show how automatically-generated weak labels can be effectively used to overcome this problem and train powerful deep learning models for labeling seismic structures in large seismic volumes. To achieve this, we automatically generate thousands of weak labels and use them to train a deconvolutional network for labeling fault, salt dome, and chaotic regions within the Netherlands F3 block. Furthermore, we show how modifying the loss function to take into account the weak training labels helps reduce false positives in the labeling results. The benefit of this work is that it enables the effective training and deployment of deep learning models to various seismic interpretation tasks without requiring any manual labeling effort. We show excellent results on the Netherlands F3 block, and show how our model outperforms other baseline models.

physics.geo-ph↗

arXiv2018

Interpretation of the diffuse astrophysical neutrino flux in terms of the blazar sequence

We study if the diffuse astrophysical neutrino flux can come from blazar jets -- a subclass of Active Galactic Nuclei (AGNs) -- while it, at the same time, respects the blazar stacking limit based on source catalogs and is consistent with the observation from TXS 0506+056. We compute the neutrino flux from resolved and unresolved sources using an averaged, empirical relationship between electromagnetic spectrum and luminosity, known as the {\em blazar sequence}, for two populations of blazars (BL Lacs and FSRQs). Using a source model with realistic neutrino flux computations, we demonstrate that blazars can indeed power the diffuse neutrino flux at the highest energies and obey the stacking limit at the same time, and we derive the conditions for the baryonic loading (proton versus $γ$-ray luminosity) evolving over the blazar sequence. Under the hypothesis that low-luminosity blazars power the diffuse astrophysical neutrino flux, we find that the dominant contribution of the diffuse flux up to PeV energies must come from unresolved BL Lacs with baryonic loadings larger than about $10^5$ -- while only a very small contribution may come from resolved high-luminosity BL Lacs or FSRQs, which can be directly tested by the stacking limit. We find that the blazar TXS 0506+056 is on the verge of these populations in our baseline scenario, at a relatively high luminosity and redshift; as a consequence we predict about 0.3 $γ$-ray-neutrino associations per year from the whole population, dominated by BL Lacs with $L_γ\simeq 10^{45} \, \mathrm{erg/s}$ and $z \sim 0.1$.

astro-ph.HE↗

arXiv2018

The $g$-good neighbor conditional diagnosability of locally exchanged twisted cubes

Connectivity and diagnosability are important parameters in measuring the fault tolerance and reliability of interconnection networks. The $R^g$-vertex-connectivity of a connected graph $G$ is the minimum cardinality of a faulty set $X\subseteq V(G)$ such that $G-X$ is disconnected and every fault-free vertex has at least $g$ fault-free neighbors. The $g$-good-neighbor conditional diagnosability is defined as the maximum cardinality of a $g$-good-neighbor conditional faulty set that the system can guarantee to identify. The interconnection network considered here is the locally exchanged twisted cube $LeTQ(s,t)$. For $1\leq s\leq t$ and $0\leq g\leq s$, we first determine the $R^g$-vertex-connectivity of $LeTQ(s,t)$, then establish the $g$-good neighbor conditional diagnosability of $LeTQ(s,t)$ under the PMC model and MM$^*$ model, respectively.

math.CO↗

arXiv2018

Neutrinos and Ultra-High-Energy Cosmic-Ray Nuclei from Blazars

We discuss the production of ultra-high-energy cosmic ray (UHECR) nuclei and neutrinos from blazars. We compute the nuclear cascade in the jet for both BL Lac objects and flat-spectrum radio quasars (FSRQs), and in the ambient radiation zones for FSRQs as well. By modeling representative spectral energy distributions along the blazar sequence, two distinct regimes are identified, which we call "nuclear survival" -- typically found in low-luminosity BL Lacs, and "nuclear cascade" -- typically found in high-luminosity FSRQs. We quantify how the neutrino and cosmic-ray (CR) emission efficiencies evolve over the blazar sequence, and demonstrate that neutrinos and CRs come from very different object classes. For example, high-frequency peaked BL Lacs (HBLs) tend to produce CRs, and HL-FSRQs are the more efficient neutrino emitters. This conclusion does not depend on the CR escape mechanism, for which we discuss two alternatives (diffusive and advective escape). Finally, the neutrino spectrum from blazars is shown to significantly depend on the injection composition into the jet, especially in the nuclear cascade case: Injection compositions heavier than protons lead to reduced neutrino production at the peak, which moves at the same time to lower energies. Thus, these sources will exhibit better compatibility with the observed IceCube and UHECR data.

astro-ph.HE↗

arXiv2017

A Graphical Social Topology Model for Multi-Object Tracking

Tracking multiple objects is a challenging task when objects move in groups and occlude each other. Existing methods have investigated the problems of group division and group energy-minimization; however, lacking overall object-group topology modeling limits their ability in handling complex object and group dynamics. Inspired with the social affinity property of moving objects, we propose a Graphical Social Topology (GST) model, which estimates the group dynamics by jointly modeling the group structure and the states of objects using a topological representation. With such topology representation, moving objects are not only assigned to groups, but also dynamically connected with each other, which enables in-group individuals to be correctly associated and the cohesion of each group to be precisely modeled. Using well-designed topology learning modules and topology training, we infer the birth/death and merging/splitting of dynamic groups. With the GST model, the proposed multi-object tracker can naturally facilitate the occlusion problem by treating the occluded object and other in-group members as a whole while leveraging overall state transition. Experiments on both RGB and RGB-D datasets confirm that the proposed multi-object tracker improves the state-of-the-arts especially in crowded scenes.

cs.CV↗