Searcharxiv⌕ Search

arXiv subjects

Cong Zhang

Publications and source records attributed to Cong Zhang.

At least 127 records · Page 7Linked to original sources
arXiv2023

Black hole image encoding quantum gravity information

The quantum extension of the Kruskal spacetime indicates the existence of a companion black hole in the universe earlier than ours. It is shown that the radiations from the companion black hole can enter its horizon, pass through the deep Planck region, and show up from the white hole in our universe. These radiations inlay extra bright rings in the image of the black hole in our universe, and some of these rings appear distinctly in the shadow region. Therefore, the image of the black hole observed by us encodes the information of quantum gravity. The positions and widths of the bright rings are predicted precisely. The predictive values for supermassive black holes are universal for a quite general class of quantum-modified spacetimes with the phenomenon of black hole to white hole transition. Thus, our result opens a new experimental window to test this phenomenon predicted by quantum gravity.

gr-qc↗
arXiv2023

Deep Progressive Feature Aggregation Network for High Dynamic Range Imaging

High dynamic range (HDR) imaging is an important task in image processing that aims to generate well-exposed images in scenes with varying illumination. Although existing multi-exposure fusion methods have achieved impressive results, generating high-quality HDR images in dynamic scenes is still difficult. The primary challenges are ghosting artifacts caused by object motion between low dynamic range images and distorted content in under and overexposed regions. In this paper, we propose a deep progressive feature aggregation network for improving HDR imaging quality in dynamic scenes. To address the issues of object motion, our method implicitly samples high-correspondence features and aggregates them in a coarse-to-fine manner for alignment. In addition, our method adopts a densely connected network structure based on the discrete wavelet transform, which aims to decompose the input features into multiple frequency subbands and adaptively restore corrupted contents. Experiments show that our proposed method can achieve state-of-the-art performance under different scenes, compared to other promising HDR imaging methods. Specifically, the HDR images generated by our method contain cleaner and more detailed content, with fewer distortions, leading to better visual quality.

cs.CV↗
arXiv2023

Light Ring behind Wormhole Throat: Geodesics, Images and Shadows

The geodesics of the Ellis-Bronnikov wormhole with two parameters are studied. The asymmetric wormhole has only one light ring and one innermost stable circular orbit located on one side of the wormhole throat. Consequently, certain light rays can be reflected back by the wormhole. Additionally, the same wormhole can have different appearances on both sides of the throat. We present novel images of the wormhole with a light ring behind the throat in a scenario with an accretion disk as the light source and in a backlit wormhole scenario, which are distinct from the images of other compact objects and have the potential to be observed.

gr-qc↗
arXiv2023

Quantum Oppenheimer-Snyder and Swiss Cheese models

By considering the quantum Oppenheimer-Snyder model in loop quantum cosmology, a new quantum black hole model whose metric tensor is a suitably deformed Schwarzschild one is derived. The quantum effects imply a lower bound on the mass of the black hole produced by the collapsing dust ball. For the case of larger masses where the event horizon does form, the maximal extension of the spacetime and its properties are investigated. By discussing the opposite scenario to the quantum Oppenheimer-Snyder, a quantum Swiss Cheese model is obtained with a bubble surrounded by the quantum universe. This model is analogous to black hole cosmology or fecund universes where the big bang is related to a white hole. Thus our models open a new window to cosmological phenomenology.

gr-qc↗
arXiv2023

Alternative $k=-1$ loop quantum cosmology

An alternative quantization of the gravitational Hamiltonian constraint of the $k=-1$ Friedmann-Robertson-Walker model is proposed by treating the Euclidean term and the Lorentzian term independently, mimicking the treatment of full loop quantum gravity. The resulting Hamiltonian constraint operator for the $k=-1$ model with a massless scalar field is successfully constructed, and is shown to have the corrected classical limit. Compared to the former quantization schemes in the literature where only the Euclidean term is quantized, the new quantum dynamics of the $k=-1$ model with a massless scalar field indicates that the classical big-bang singularity is replaced by an asymmetric quantum bounce.

gr-qc↗
arXiv2023

HierCat: Hierarchical Query Categorization from Weakly Supervised Data at Facebook Marketplace

Query categorization at customer-to-customer e-commerce platforms like Facebook Marketplace is challenging due to the vagueness of search intent, noise in real-world data, and imbalanced training data across languages. Its deployment also needs to consider challenges in scalability and downstream integration in order to translate modeling advances into better search result relevance. In this paper we present HierCat, the query categorization system at Facebook Marketplace. HierCat addresses these challenges by leveraging multi-task pre-training of dual-encoder architectures with a hierarchical inference step to effectively learn from weakly supervised training data mined from searcher engagement. We show that HierCat not only outperforms popular methods in offline experiments, but also leads to 1.4% improvement in NDCG and 4.3% increase in searcher engagement at Facebook Marketplace Search in online A/B testing.

cs.IR↗
arXiv2023

Investigating the collinear splitting effects of boosted dark matter at neutrino detectors

We study the probing prospects of cosmic ray boosted dark matter (DM) in the framework of simplified electron-philic dark photon model. Focusing on the dark matter and dark photon masses around keV $\sim $ MeV scale, we consider the bounds obtained from the XENON1T and Super-K experiments. The electron bound state effects are treated carefully in calculating the XENON1T constraint. As for the detection at neutrino detector where the energy threshold is relatively higher, the large logarithmic effects induced by the scale hierarchy between the masses and momentum transfer are considered by introducing the DM parton distribution function (PDF). The logarithmic effects will reduce the electron recoil rate for DM scattering in neutrino detectors. Moreover, we find the DUNE and JUNO experiments provide high sensitivities for probing the dark photon component in the DM PDF through the dark Compton process. We also check the Bullet Cluster constraint on the DM self-scattering cross section.

hep-ph↗
arXiv2023

Weighted EF1 Allocations for Indivisible Chores

We study how to fairly allocate a set of indivisible chores to a group of agents, where each agent $i$ has a non-negative weight $w_i$ that represents its obligation for undertaking the chores. We consider the fairness notion of weighted envy-freeness up to one item (WEF1) and propose an efficient picking sequence algorithm for computing WEF1 allocations. Our analysis is based on a natural and powerful continuous interpretation for the picking sequence algorithms in the weighted setting, which might be of independent interest. Using this interpretation, we establish the necessary and sufficient conditions under which picking sequence algorithms can guarantee other fairness notions in the weighted setting. We also study the existence of fair and efficient allocations and propose efficient algorithms for the computation of WEF1 and PO allocations for the bi-valued instances. Our result generalizes that of Garg et al. (AAAI 2022) and Ebadian et al. (AAMAS 2022) to the weighted setting. Our work also studies the price of fairness for WEF1, and the implications of WEF1 to other fairness notions.

cs.GT↗
arXiv2022

Fermions in Loop Quantum Gravity and Resolution of Doubling Problem

The fermion propagator is derived in detail from the model of fermion coupled to loop quantum gravity. As an ingredient of the propagator, the vacuum state is defined as the ground state of some effective fermion Hamiltonian under the background geometry given by a coherent state resembling the classical Minkowski spacetime. Moreover, as a critical feature of loop quantum gravity, the superposition over graphs is employed to define the vacuum state. It turns out that the graph superposition leads to the propagator being the average of the propagators of the lattice field theory over various graphs so that all fermion doubler modes are suppressed in the propagator. This resolves the doubling problem in loop quantum gravity. Our result suggests that the superposition nature of quantum geometry should, on the one hand, resolve the tension between fermion and the fundamental discreteness and, on the other hand, relate to the continuum limit of quantum gravity.

gr-qc↗
arXiv2022

Bootstrapping meaning through listening: Unsupervised learning of spoken sentence embeddings

Inducing semantic representations directly from speech signals is a highly challenging task but has many useful applications in speech mining and spoken language understanding. This study tackles the unsupervised learning of semantic representations for spoken utterances. Through converting speech signals into hidden units generated from acoustic unit discovery, we propose WavEmbed, a multimodal sequential autoencoder that predicts hidden units from a dense representation of speech. Secondly, we also propose S-HuBERT to induce meaning through knowledge distillation, in which a sentence embedding model is first trained on hidden units and passes its knowledge to a speech encoder through contrastive learning. The best performing model achieves a moderate correlation (0.5~0.6) with human judgments, without relying on any labels or transcriptions. Furthermore, these models can also be easily extended to leverage textual transcriptions of speech to learn much better speech embeddings that are strongly correlated with human annotations. Our proposed methods are applicable to the development of purely data-driven systems for speech mining, indexing and search.

cs.CL↗
arXiv2022

Learning to Solve Multiple-TSP with Time Window and Rejections via Deep Reinforcement Learning

We propose a manager-worker framework based on deep reinforcement learning to tackle a hard yet nontrivial variant of Travelling Salesman Problem (TSP), \ie~multiple-vehicle TSP with time window and rejections (mTSPTWR), where customers who cannot be served before the deadline are subject to rejections. Particularly, in the proposed framework, a manager agent learns to divide mTSPTWR into sub-routing tasks by assigning customers to each vehicle via a Graph Isomorphism Network (GIN) based policy network. A worker agent learns to solve sub-routing tasks by minimizing the cost in terms of both tour length and rejection rate for each vehicle, the maximum of which is then fed back to the manager agent to learn better assignments. Experimental results demonstrate that the proposed framework outperforms strong baselines in terms of higher solution quality and shorter computation time. More importantly, the trained agents also achieve competitive performance for solving unseen larger instances.

cs.LG↗
arXiv2022

ByT5 model for massively multilingual grapheme-to-phoneme conversion

In this study, we tackle massively multilingual grapheme-to-phoneme conversion through implementing G2P models based on ByT5. We have curated a G2P dataset from various sources that covers around 100 languages and trained large-scale multilingual G2P models based on ByT5. We found that ByT5 operating on byte-level inputs significantly outperformed the token-based mT5 model in terms of multilingual G2P. Pairwise comparison with monolingual models in these languages suggests that multilingual ByT5 models generally lower the phone error rate by jointly learning from a variety of languages. The pretrained model can further benefit low resource G2P through zero-shot prediction on unseen languages or provides pretrained weights for finetuning, which helps the model converge to a lower phone error rate than randomly initialized weights. To facilitate future research on multilingual G2P, we make available our code and pretrained multilingual G2P models at: https://github.com/lingjzhu/CharsiuG2P.

cs.CL↗
arXiv2022

Sampling Efficient Deep Reinforcement Learning through Preference-Guided Stochastic Exploration

Massive practical works addressed by Deep Q-network (DQN) algorithm have indicated that stochastic policy, despite its simplicity, is the most frequently used exploration approach. However, most existing stochastic exploration approaches either explore new actions heuristically regardless of Q-values or inevitably introduce bias into the learning process to couple the sampling with Q-values. In this paper, we propose a novel preference-guided $ε$-greedy exploration algorithm that can efficiently learn the action distribution in line with the landscape of Q-values for DQN without introducing additional bias. Specifically, we design a dual architecture consisting of two branches, one of which is a copy of DQN, namely the Q-branch. The other branch, which we call the preference branch, learns the action preference that the DQN implicit follows. We theoretically prove that the policy improvement theorem holds for the preference-guided $ε$-greedy policy and experimentally show that the inferred action preference distribution aligns with the landscape of corresponding Q-values. Consequently, preference-guided $ε$-greedy exploration motivates the DQN agent to take diverse actions, i.e., actions with larger Q-values can be sampled more frequently whereas actions with smaller Q-values still have a chance to be explored, thus encouraging the exploration. We assess the proposed method with four well-known DQN variants in nine different environments. Extensive results confirm the superiority of our proposed method in terms of performance and convergence speed. Index Terms- Preference-guided exploration, stochastic policy, data efficiency, deep reinforcement learning, deep Q-learning.

cs.LG↗
arXiv2022

Fermion coupling to loop quantum gravity: canonical formulation

In the model of a fermion field coupled to loop quantum gravity, we consider the Gauss and the Hamiltonian constraints. According to the explicit solutions to the Gauss constraint, the fermion spins and the gravitational spin networks intertwine with each other so that the fermion spins contribute to the volume of the spin network vertices. For the Hamiltonian constraint, the regularization and quantization procedures are presented in detail. By introducing an adapted vertex Hilbert space to remove the regulator, we propose a diffeomorphism covariant graph-changing Hamiltonian constraint operator of the fermion field. This operator shows how fermions move in the loop quantum gravity spacetime and simultaneously influences the background quantum geometry.

gr-qc↗
arXiv2022

Fermions on Quantum Geometry and Resolution of Doubling Problem

The fermion doubling problem has an important impact on quantum gravity, by revealing the tension between fermion and the fundamental discreteness of quantum spacetime. In this work, we discover that in Loop Quantum Gravity, the quantum geometry involving superposition of states associated with lattice refinements provides a resolution to the fermion doubling problem. We construct and analyze the fermion propagator on the quantum geometry, and we show that all fermion doubler modes are suppressed in the propagator. Our result suggests that the superposition nature of quantum geometry should resolve the tension between fermion and the fundamental discreteness, and relate to the continuum limit of quantum gravity.

gr-qc↗
arXiv2022

First-Order Quantum Correction in Coherent State Expectation Value of Loop-Quantum-Gravity Hamiltonian: Overview and Results

Given the Loop-Quantum-Gravity (LQG) non-graph-changing Hamiltonian $\widehat{H[N]}$, the coherent state expectation value $\langle\widehat{H[N]}\rangle$ admits an semiclassical expansion in $\ell^2_{\rm p}$. In this paper, we compute explicitly the expansion of $\langle\widehat{H[N]}\rangle$ on the cubic graph to the linear order in $\ell^2_{\rm p}$, when the coherent state is peaked at the homogeneous and isotropic data of cosmology. In our computation, a powerful algorithm is developed to overcome the complexity in computing $\langle \widehat{H[N]} \rangle$. In particular, some key innovations in our algorithm substantially reduce the computational complexity in the Lorentzian part of $\langle\widehat{H[N]}\rangle$. Moreover, the algorithm developed in the present work makes it possible to compute the expectation value of arbitrary monomial of holonomies and fluxes on one edge up to arbitrary order of $\ell_{\rm p}^2$.

gr-qc↗
arXiv2022

VSEGAN: Visual Speech Enhancement Generative Adversarial Network

Speech enhancement is an essential task of improving speech quality in noise scenario. Several state-of-the-art approaches have introduced visual information for speech enhancement,since the visual aspect of speech is essentially unaffected by acoustic environment. This paper proposes a novel frameworkthat involves visual information for speech enhancement, by in-corporating a Generative Adversarial Network (GAN). In par-ticular, the proposed visual speech enhancement GAN consistof two networks trained in adversarial manner, i) a generator that adopts multi-layer feature fusion convolution network to enhance input noisy speech, and ii) a discriminator that attemptsto minimize the discrepancy between the distributions of the clean speech signal and enhanced speech signal. Experiment re-sults demonstrated superior performance of the proposed modelagainst several state-of-the-art

eess.AS↗