arXiv subjects

Bin Zhu

Publications and source records attributed to Bin Zhu.

At least 199 records · Page 11Linked to original sources

arXiv2022

EPIC-KITCHENS VISOR Benchmark: VIdeo Segmentations and Object Relations

We introduce VISOR, a new dataset of pixel annotations and a benchmark suite for segmenting hands and active objects in egocentric video. VISOR annotates videos from EPIC-KITCHENS, which comes with a new set of challenges not encountered in current video segmentation datasets. Specifically, we need to ensure both short- and long-term consistency of pixel-level annotations as objects undergo transformative interactions, e.g. an onion is peeled, diced and cooked - where we aim to obtain accurate pixel-level annotations of the peel, onion pieces, chopping board, knife, pan, as well as the acting hands. VISOR introduces an annotation pipeline, AI-powered in parts, for scalability and quality. In total, we publicly release 272K manual semantic masks of 257 object classes, 9.9M interpolated dense masks, 67K hand-object relations, covering 36 hours of 179 untrimmed videos. Along with the annotations, we introduce three challenges in video object segmentation, interaction understanding and long-term reasoning. For data, code and leaderboards: http://epic-kitchens.github.io/VISOR

cs.CV↗

arXiv2022

Probing Supersymmetry Breaking Scale with Atomic Clocks

The supersymmetry (SUSY) breaking mechanism generally predicts the existence of the sgoldstinos, which can play the role of wave-like dark matter. Due to the ubiquitous coupling to the electromagnetic fields, the light scalar sgoldstino dark matter can lead to the variance of the fine-structure constant. With the precise atomic clock data, we find the SUSY breaking scale $\sqrt{F}$ can be probed up to the GUT scale in the sgoldstino mass range of $10^{-22}$ eV $<m_ϕ<$ $4\times10^{-7}$ eV.

hep-ph↗

arXiv2022

Celestial Liouville Theory for Yang-Mills Amplitudes

We consider Yang-Mills theory with the coupling constant and theta angle determined by the vacuum expectation values of a dynamical (complex) dilaton field. We discuss the tree-level N-gluon MHV scattering amplitudes in the presence of a nontrivial background dilaton field and construct the corresponding celestial amplitudes by taking Mellin transforms with respect to the lightcone energies. In this way, we obtain two-dimensional CFT correlators of primary fields on the celestial sphere. We show that the celestial Yang-Mills amplitudes evaluated in the presence of a spherical dilaton shockwave are given by the correlation functions of primary field operators factorized into the holomorphic current operators times the "light" Liouville operators. They are evaluated in the semiclassical limit of Liouville theory (the limit of infinite central charge) and are determined by the classical Liouville field describing metrics on the celestial sphere.

hep-th↗

arXiv2022

Sampling Gaussian Stationary Random Fields: A Stochastic Realization Approach

Generating large-scale samples of stationary random fields is of great importance in the fields such as geomaterial modeling and uncertainty quantification. Traditional methodologies based on covariance matrix decomposition have the diffculty of being computationally expensive, which is even more serious when the dimension of the random field is large. This paper proposes an effcient stochastic realization approach for sampling Gaussian stationary random fields from a systems and control point of view. Specifically, we take the exponential and Gaussian covariance functions as examples and make a decoupling assumption when there are multiple dimensions. Then a rational spectral density is constructed in each dimension using techniques from covariance extension, and the corresponding autoregressive moving-average (ARMA) model is obtained via spectral factorization. As a result, samples of the random field with a specific covariance function can be generated very effciently in the space domain by implementing the ARMA recursion using a white noise input. Such a procedure is computationally cheap due to the fact that the constructed ARMA model has a low order. Furthermore, the same method is integrated to multiscale simulations where interpolations of the generated samples are achieved when one zooms into finer scales. Both theoretical analysis and simulation results show that our approach performs favorably compared with covariance matrix decomposition methods.

stat.ME↗

arXiv2022

Correlating Gravitational Waves with $W$-boson Mass, FIMP Dark Matter, and Majorana Seesaw Mechanism

We study a minimal extension of the Standard Model by introducing three right-handed neutrinos and a new scotogenic scalar doublet, in which the mass splittings between neutral and charged components are responsible for the $W$-boson mass newly measured by the CDF collaboration. This model can not only generate non-vanishing Majorana neutrino masses via the interaction of right-handed neutrinos and scotogenic scalars, but also explain the Universe's missing matter in the form of FIMP dark matter. We also study the influence of the mass splitting on the first order electroweak phase transition, and find that it can further enhance the transition strength and thus induce gravitational waves during the phase transition, which may be detected in the forthcoming detectors such as U-DECIGO.

hep-ph↗

arXiv2022

Celestial Yang-Mills Amplitudes and D=4 Conformal Blocks

We discuss the properties of recently constructed "single-valued" celestial four-gluon amplitudes. We show that the amplitude factorizes into the "current" part and the "scalar" part. The current factor is given by the group-dependent part of the Wess-Zumino-Witten correlator of four holomorphic currents with a non-vanishing level of Kač-Moody algebra. The scalar factor can be expressed in terms of a complex integral of the Koba-Nielsen form, similar to the integrals describing four-point correlators in Coulomb gas models and, more generally, in the infinite central charge limit of Liouville theory. The scalar part can be also obtained by a dimensional reduction of a single D=4 conformal block and the shadow block from Minkowski space to the celestial sphere.

hep-th↗

arXiv2022

Cross-lingual Adaptation for Recipe Retrieval with Mixup

Cross-modal recipe retrieval has attracted research attention in recent years, thanks to the availability of large-scale paired data for training. Nevertheless, obtaining adequate recipe-image pairs covering the majority of cuisines for supervised learning is difficult if not impossible. By transferring knowledge learnt from a data-rich cuisine to a data-scarce cuisine, domain adaptation sheds light on this practical problem. Nevertheless, existing works assume recipes in source and target domains are mostly originated from the same cuisine and written in the same language. This paper studies unsupervised domain adaptation for image-to-recipe retrieval, where recipes in source and target domains are in different languages. Moreover, only recipes are available for training in the target domain. A novel recipe mixup method is proposed to learn transferable embedding features between the two domains. Specifically, recipe mixup produces mixed recipes to form an intermediate domain by discretely exchanging the section(s) between source and target recipes. To bridge the domain gap, recipe mixup loss is proposed to enforce the intermediate domain to locate in the shortest geodesic path between source and target domains in the recipe embedding space. By using Recipe 1M dataset as source domain (English) and Vireo-FoodTransfer dataset as target domain (Chinese), empirical experiments verify the effectiveness of recipe mixup for cross-lingual adaptation in the context of image-to-recipe retrieval.

cs.CV↗

arXiv2022

Models for Self-resonant Dark Matter

We consider a new mechanism for enhancing the self-scattering and annihilation cross sections for dark matter with multiple components but without a light mediator. The lighter dark matter component plays a role of the $u$-channel pole in the elastic co-scattering for dark matter, leading to a large self-scattering cross section and a Sommerfeld enhancement for semi-annihilation processes. Taking the effective theory approach for self-resonant dark matter, we present various combinations of multiple dark matter components with spins and parities, showing a $u$-channel pole in the co-scattering processes. Adopting dark photon and dark Higgs portals for self-resonant dark matter, we impose the relic density condition as well as indirect detection bounds on semi-annihilation channels with a Sommerfeld enhancement and discuss potential signals for direct detection experiments.

hep-ph↗

arXiv2022

Improving robustness of language models from a geometry-aware perspective

Recent studies have found that removing the norm-bounded projection and increasing search steps in adversarial training can significantly improve robustness. However, we observe that a too large number of search steps can hurt accuracy. We aim to obtain strong robustness efficiently using fewer steps. Through a toy experiment, we find that perturbing the clean data to the decision boundary but not crossing it does not degrade the test accuracy. Inspired by this, we propose friendly adversarial data augmentation (FADA) to generate friendly adversarial data. On top of FADA, we propose geometry-aware adversarial training (GAT) to perform adversarial training on friendly adversarial data so that we can save a large number of search steps. Comprehensive experiments across two widely used datasets and three pre-trained language models demonstrate that GAT can obtain stronger robustness via fewer steps. In addition, we provide extensive empirical results and in-depth analyses on robustness to facilitate future studies.

cs.CL↗

arXiv2022

Axion Dark Radiation: Hubble Tension and Hyper-kamiokande Neutrino Experiment

In this work, we investigate the dark sector of a supersymmetric axion model, consisting of the late-decaying gravitino/axino dark matter and axion dark radiation. In the early universe, the decay of the scalar superpartner of the axion (saxion) will produce a large amount of entropy. The additional entropy can not only dilute the relic density of the gravitino/axino dark matter to avoid overclosing the universe but also relax the constraint on the reheating temperature $T_{R}$ after inflation. Meanwhile, the axion dark radiation from the saxion decay will increase the effective number of neutrino species $N_{\rm eff}$, which can help to reduce the cosmological Hubble tension. In the late universe, the decay of long-lived gravitino/axino dark matter produces the axions with MeV-scale kinetic energy. We study the potential of searching for such energetic axions through the inverse Primakoff process $a+A \to γ+ A$ in the neutrino experiments, such as Hyper-Kamiokande.

hep-ph↗

arXiv2022

Electroweak Precision Fit and New Physics in light of $W$ Boson Mass

The $W$ boson mass is one of the most important electroweak precision observables for testing the Standard Model or its extensions. The very recent measured $W$ boson mass at CDF shows about $7σ$ deviations from the SM prediction, which may challenge the internal consistency of the SM. By performing the global electroweak fit with the new $W$-boson, we present the new values of the oblique parameters: $S = 0.06 \pm 0.10$, $T= 0.11 \pm 0.12$, $U=0.13 \pm 0.09$, or $S=0.14 \pm 0.08$, $T= 0.26 \pm 0.06$ with $U =0$ and the corresponding correlation matrices, which strongly indicates the need for the non-degenerate multiplets beyond the SM. As a proof-of-concept, we show that the new results can be accommodated in the two-Higgs doublet model, where the charged Higgs boson has to be either heavier or lighter than both two heavy neutral Higgs bosons. Therefore, searching for these non-SM Higgs bosons will provide a complementary way to test the new physics for the $W$ boson mass anomaly.

hep-ph↗

arXiv2022

The Realistic Scattering of Puffy Dark Matter

If dark matter has a finite size, the intrinsic interaction responsible for the structure formation is inevitable from the perspective of dark matter self-scattering. To describe the circumstance in which the binding force realizes the finite size dark protons, we first use the Eikonal approximation to simplify the convoluted scattering between dark protons into the case at the $t=0$ limit. The Chou-Yang model is then introduced to reduce the number of input parameters to one based on the simplicity and analyticity principle. A new definition of velocity dependence and the corresponding implications on the small cosmological structures from Chou-Yang dark protons are shown clearly. Even though the parameter space is not fully covered, the numerical findings show that the amplitude coefficient can alter the self-scattering cross-section, allowing us to recover the excluded parameter space without using binding force. Finally, we demonstrate that the correct relic density from thermal freeze-out production prefers super heavy dark protons.

hep-ph↗

arXiv2022

OneLabeler: A Flexible System for Building Data Labeling Tools

Labeled datasets are essential for supervised machine learning. Various data labeling tools have been built to collect labels in different usage scenarios. However, developing labeling tools is time-consuming, costly, and expertise-demanding on software development. In this paper, we propose a conceptual framework for data labeling and OneLabeler based on the conceptual framework to support easy building of labeling tools for diverse usage scenarios. The framework consists of common modules and states in labeling tools summarized through coding of existing tools. OneLabeler supports configuration and composition of common software modules through visual programming to build data labeling tools. A module can be a human, machine, or mixed computation procedure in data labeling. We demonstrate the expressiveness and utility of the system through ten example labeling tools built with OneLabeler. A user study with developers provides evidence that OneLabeler supports efficient building of diverse data labeling tools.

cs.HC↗

arXiv2022

Probing ultra-light dark photon from inverse Compton-like scattering

Dark photon not only provides a portal linking dark sector particles and ordinary matter but also is a well-motivated dark matter candidate. We propose to detect the dark photon dark matter through the inverse Compton-like scattering process $p+γ^\prime \to p+γ$. Thanks to the ultra-high energy primary cosmic rays, we find that such a method is able to probe the dark photon mass from $10^{-2}$ eV down to $10^{-19}$ eV with the expected sensitivity of eROSITA $X$-ray telescope, which can extend the current lower limit of dark photon mass from Jupiter's magnetic fields experiment by about three orders of magnitude.

hep-ph↗

arXiv2022

Elements of Celestial Conformal Field Theory

In celestial holography, four-dimensional scattering amplitudes are considered as two-dimensional conformal correlators of a putative two-dimensional celestial conformal field theory (CCFT). The simplest way of converting momentum space amplitudes into CCFT correlators is by taking their Mellin transforms with respect to light-cone energies. For massless particles, like gluons, however, such a construction leads to three-point and four-point correlators that vanish everywhere except for a measure zero hypersurface of celestial coordinates. This is due to the four-dimensional momentum conservation law that constrains the insertion points of the operators associated with massless particles. These correlators are reminiscent of Coulomb gas correlators that, in the absence of background charges, vanish due to charge conservation. We supply the background momentum by coupling Yang-Mills theory to a background dilaton field, with the (complex) dilaton source localized on the celestial sphere. This picture emerges from the physical interpretation of the solutions of the system of differential equations discovered by Banerjee and Ghosh. We show that the solutions can be written as Mellin transforms of the amplitudes evaluated in such a dilaton background. The resultant three-gluon and four-gluon amplitudes are single-valued functions of celestial coordinates enjoying crossing symmetry and all other properties expected from standard CFT correlators. We use them to extract OPEs and compare them with the OPEs extracted from multi-gluon celestial amplitudes without a dilaton background. We perform the conformal block decomposition of the four-gluon single-valued correlator and determine the dimensions, spin and group representations of the entire primary field spectrum of the Yang-Mills sector of CCFT.

hep-th↗

arXiv2022

Low energy supersymmetry confronted with current experiments: an overview

This study provides a brief overview of low-energy supersymmetry (SUSY) in light of current experimental constraints, such as collider searches, dark matter searches, and muon $g-2$ measurements. In addition, we survey a variety of low energy supersymmetric models: the phenomenological minimal supersymmetric model (MSSM); the supersymmetric models with cut-off-scale boundary conditions, i.e., the minimal supergravity (mSUGRA) or the constrained MSSM (CMSSM), the gauge mediation of SUSY breaking (GMSB), and the anomaly mediation of SUSY breaking (AMSB), as well as their extensions. The conclusion is that the low energy SUSY can survive all current experimental constraints and remains compelling, albeit suffering from a little fine-tuning problem. The fancy models like mSUGRA, GMSB, and AMSB need to be extended if the muon $g-2$ anomaly comes from new physics.

hep-ph↗

arXiv2021

Probing the flavor-specific scalar mediator for the muon $(g-2)$ deviation, the proton radius puzzle and the light dark matter production

Flavor-specific scalar bosons exist in various Standard Model extensions and couple to a single generation of fermions via a global flavor symmetry breaking mechanism. Given this strategy, we propose a MeV flavor-specific scalar model in dimension-$5$ operator series, which explains the muon g-2 anomaly and proton radius puzzle by coupling with the muon and down-quark at the same time. The framework is consistent with the null result of high-intensity searches. Specifically, the supernova constraints for muon couplings become weakened by including the contribution of down-quark interaction. The parameter space for explaining muon $g-2$ discrepancy is available when $10\%$ energy deposition is required in the energy explosion process in the supernova, but this is ruled out by the $1\%$ energy deposition requirement. We also investigate the searches for mediator and dark matter and the resulting constraints on viable parameter space such as nuclear physics constraints, direct detection for light boosted dark matter, and possible CMB constraints. When compared to conventional dark matter production, light dark matter production has two additional modifications: bound state formation and early kinetic equilibrium decoupling. We are now looking into the implications of these effects on the relic density of light dark matter.

hep-ph↗

arXiv2021

Direct Detection of Spin-Dependent Sub-GeV Dark Matter via Migdal Effect

Motivated by the current strong constraints on the spin-independent dark matter (DM)-nucleus scattering, we investigate the spin-dependent (SD) interactions of the light Majorana DM with the nucleus mediated by an axial-vector boson. Due to the small nucleus recoil energy, the ionization signals have now been used to probe the light dark matter particles in direct detection experiments. With the existing ionization data, we derive the exclusion limits on the SD DM-nucleus scattering through Migdal effect in the MeV-GeV DM mass range. It is found that the lower limit of the DM mass can reach about several MeVs. Due to the momentum transfer correction induced by the light mediator, the bounds on the SD DM-nucleus scattering cross sections can be weakened in comparison with the heavy mediator.

hep-ph↗