arXiv subjects

Hao Yang

Publications and source records attributed to Hao Yang.

At least 343 records · Page 19Linked to original sources

arXiv2022

Internal structure and its connection with thermodynamics and dynamics in black holes

In general, a finite metric function at the center of a black hole describes a non-singular spacetime but an infinite metric at the center gives a singular spacetime, where the former is associated with convergent Ricci and Kretschmann scalars together with complete radial geodesics, while the latter is related to divergent Ricci and Kretschmann scalars together with incomplete radial geodesics. For the charged black hole in the four-dimensional Einstein-Gauss-Bonnet theory, we find that its metric function is finite at its center in one region of parameters but complex in the other region of parameters. The finite case describes a strange spacetime which presents the Ricci and Kretschmann scalars of a singular spacetime and the radial geodesics of a non-singular spacetime, while the complex case gives rise to the similar situation. We verify that the cosmic censorship conjecture is maintained for the black hole model. Further, we investigate the second-order phase transition, quasinormal modes of perturbation of a test scalar field in the eikonal limit, and shadow radius for the black hole and find that the thermodynamic and dynamic properties depend on the metrics. In this way, we connect the internal structure with the thermodynamics and dynamics for this black hole. Moreover, we compare such a black hole of modified gravity with the Reissner-Nordström black hole of Einstein's general relativity in these thermodynamic and dynamic properties.

gr-qc↗

arXiv2022

ComplETR: Reducing the cost of annotations for object detection in dense scenes with vision transformers

Annotating bounding boxes for object detection is expensive, time-consuming, and error-prone. In this work, we propose a DETR based framework called ComplETR that is designed to explicitly complete missing annotations in partially annotated dense scene datasets. This reduces the need to annotate every object instance in the scene thereby reducing annotation cost. ComplETR augments object queries in DETR decoder with patch information of objects in the image. Combined with a matching loss, it can effectively find objects that are similar to the input patch and complete the missing annotations. We show that our framework outperforms the state-of-the-art methods such as Soft Sampling and Unbiased Teacher by itself, while at the same time can be used in conjunction with these methods to further improve their performance. Our framework is also agnostic to the choice of the downstream object detectors; we show performance improvement for several popular detectors such as Faster R-CNN, Cascade R-CNN, CenterNet2, and Deformable DETR on multiple dense scene datasets.

cs.CV↗

arXiv2022

Accessibility of SPDEs driven by pure jump noise and its applications

In this paper, we develop a new method to obtain the accessibility of stochastic partial differential equations driven by additive pure jump noise. An important novelty of this paper is to allow the driving noises to be degenerate. As an application, for the first time, we obtain the accessibility of a class of stochastic equations driven by pure jump degenerate noise, which cover 2D stochastic Navier-Stokes equations, stochastic Burgers type equations, singular stochastic p-Laplace equations, stochastic fast diffusion equations, etc. As a further application, we establish the ergodicity of singular stochastic p-Laplace equations and stochastic fast diffusion equations driven by additive pure jump noise, and we remark that the driving noises could be Levy processes with heavy tails.

math.PR↗

arXiv2022

LogStamp: Automatic Online Log Parsing Based on Sequence Labelling

Logs are one of the most critical data for service management. It contains rich runtime information for both services and users. Since size of logs are often enormous in size and have free handwritten constructions, a typical log-based analysis needs to parse logs into structured format first. However, we observe that most existing log parsing methods cannot parse logs online, which is essential for online services. In this paper, we present an automatic online log parsing method, name as LogStamp. We extensively evaluate LogStamp on five public datasets to demonstrate the effectiveness of our proposed method. The experiments show that our proposed method can achieve high accuracy with only a small portion of the training set. For example, it can achieve an average accuracy of 0.956 when using only 10% of the data training.

cs.SE↗

arXiv2022

Prompt Tuning for Generative Multimodal Pretrained Models

Prompt tuning has become a new paradigm for model tuning and it has demonstrated success in natural language pretraining and even vision pretraining. In this work, we explore the transfer of prompt tuning to multimodal pretraining, with a focus on generative multimodal pretrained models, instead of contrastive ones. Specifically, we implement prompt tuning on the unified sequence-to-sequence pretrained model adaptive to both understanding and generation tasks. Experimental results demonstrate that the light-weight prompt tuning can achieve comparable performance with finetuning and surpass other light-weight tuning methods. Besides, in comparison with finetuned models, the prompt-tuned models demonstrate improved robustness against adversarial attacks. We further figure out that experimental factors, including the prompt length, prompt depth, and reparameteratization, have great impacts on the model performance, and thus we empirically provide a recommendation for the setups of prompt tuning. Despite the observed advantages, we still find some limitations in prompt tuning, and we correspondingly point out the directions for future studies. Codes are available at \url{https://github.com/OFA-Sys/OFA}

cs.CL↗

arXiv2022

Rethinking Few-Shot Object Detection on a Multi-Domain Benchmark

Most existing works on few-shot object detection (FSOD) focus on a setting where both pre-training and few-shot learning datasets are from a similar domain. However, few-shot algorithms are important in multiple domains; hence evaluation needs to reflect the broad applications. We propose a Multi-dOmain Few-Shot Object Detection (MoFSOD) benchmark consisting of 10 datasets from a wide range of domains to evaluate FSOD algorithms. We comprehensively analyze the impacts of freezing layers, different architectures, and different pre-training datasets on FSOD performance. Our empirical results show several key factors that have not been explored in previous works: 1) contrary to previous belief, on a multi-domain benchmark, fine-tuning (FT) is a strong baseline for FSOD, performing on par or better than the state-of-the-art (SOTA) algorithms; 2) utilizing FT as the baseline allows us to explore multiple architectures, and we found them to have a significant impact on down-stream few-shot tasks, even with similar pre-training performances; 3) by decoupling pre-training and few-shot learning, MoFSOD allows us to explore the impact of different pre-training datasets, and the right choice can boost the performance of the down-stream tasks significantly. Based on these findings, we list possible avenues of investigation for improving FSOD performance and propose two simple modifications to existing algorithms that lead to SOTA performance on the MoFSOD benchmark. The code is available at https://github.com/amazon-research/few-shot-object-detection-benchmark.

cs.CV↗

arXiv2022

Boosting 3D Object Detection via Object-Focused Image Fusion

3D object detection has achieved remarkable progress by taking point clouds as the only input. However, point clouds often suffer from incomplete geometric structures and the lack of semantic information, which makes detectors hard to accurately classify detected objects. In this work, we focus on how to effectively utilize object-level information from images to boost the performance of point-based 3D detector. We present DeMF, a simple yet effective method to fuse image information into point features. Given a set of point features and image feature maps, DeMF adaptively aggregates image features by taking the projected 2D location of the 3D point as reference. We evaluate our method on the challenging SUN RGB-D dataset, improving state-of-the-art results by a large margin (+2.1 mAP@0.25 and +2.3mAP@0.5). Code is available at https://github.com/haoy945/DeMF.

cs.CV↗

arXiv2022

Quantum phase transition in magnetic nanographenes on a lead superconductor

Quantum spins, referred to the spin operator preserved by full SU(2) symmetry in the absence of the magnetic anistropy, have been proposed to host exotic interactions with superconductivity4. However, spin orbit coupling and crystal field splitting normally cause a significant magnetic anisotropy for d/f-shell spins on surfaces6,9, breaking SU(2) symmetry and fabricating the spins with Ising properties10. Recently, magnetic nanographenes have been proven to host intrinsic quantum magnetism due to their negligible spin orbital coupling and crystal field splitting. Here, we fabricate three atomically precise nanographenes with the same magnetic ground state of spin S=1/2 on Pb(111) through engineering sublattice imbalance in graphene honeycomb lattice. Scanning tunneling spectroscopy reveals the coexistence of magnetic bound states and Kondo screening in such hybridized system. Through engineering the magnetic exchange strength between the unpaired spin in nanographenes and cooper pairs, quantum phase transition from the singlet to the doublet state has been observed, in consistent with quantum models of spins on superconductors. Our work demonstrates delocalized graphene magnetism host highly tunable magnetic bound states with cooper pairs, which can be further developed to study the Majorana bound states and other rich quantum physics of low-dimensional quantum spins on superconductors.

cond-mat.mes-hall↗

arXiv2022

M-Adapter: Modality Adaptation for End-to-End Speech-to-Text Translation

End-to-end speech-to-text translation models are often initialized with pre-trained speech encoder and pre-trained text decoder. This leads to a significant training gap between pre-training and fine-tuning, largely due to the modality differences between speech outputs from the encoder and text inputs to the decoder. In this work, we aim to bridge the modality gap between speech and text to improve translation quality. We propose M-Adapter, a novel Transformer-based module, to adapt speech representations to text. While shrinking the speech sequence, M-Adapter produces features desired for speech-to-text translation via modelling global and local dependencies of a speech sequence. Our experimental results show that our model outperforms a strong baseline by up to 1 BLEU score on the Must-C En$\rightarrow$DE dataset.\footnote{Our code is available at https://github.com/mingzi151/w2v2-st.}

cs.CL↗

arXiv2022

MACSA: A Multimodal Aspect-Category Sentiment Analysis Dataset with Multimodal Fine-grained Aligned Annotations

Multimodal fine-grained sentiment analysis has recently attracted increasing attention due to its broad applications. However, the existing multimodal fine-grained sentiment datasets most focus on annotating the fine-grained elements in text but ignore those in images, which leads to the fine-grained elements in visual content not receiving the full attention they deserve. In this paper, we propose a new dataset, the Multimodal Aspect-Category Sentiment Analysis (MACSA) dataset, which contains more than 21K text-image pairs. The dataset provides fine-grained annotations for both textual and visual content and firstly uses the aspect category as the pivot to align the fine-grained elements between the two modalities. Based on our dataset, we propose the Multimodal ACSA task and a multimodal graph-based aligned model (MGAM), which adopts a fine-grained cross-modal fusion method. Experimental results show that our method can facilitate the baseline comparison for future research on this corpus. We will make the dataset and code publicly available.

cs.CL↗

arXiv2022

High sensitivity air-coupled MHz frequency ultrasound detection using on-chip microcavities

Owing to their dual-resonance enhanced sensitivity, cavity optomechanical systems provide an ideal platform for ultrasound sensing. In this work, we realize high sensitivity air-coupled ultrasound sensing from kilohertz (kHz) to megahertz (MHz) frequency range based on whispering gallery mode microcavities. Using a 57 um-diameter microtoroid with high optical Q factor (~10^7) and mechanical Q factor (~700), we achieve sensitivities of 46 uPa Hz^{-1/2}-10 mPa Hz^{-1/2} in a frequency range of 0.25-3.2 MHz. Thermal-noise-limited sensitivity is realized around the mechanical resonance at 2.56 MHz, in a frequency range of 0.6 MHz. We also observe the second- and third-order mechanical sidebands, and quantitatively study the intensities of each mechanical sideband as a function of the mechanical displacement. Measuring the combination of signal to noise ratios at all sidebands has the potential to extend the dynamic range of ultrasound sensing. In addition, to improve the ultrasound sensitivity in the kHz frequency range, we use a microdisk with a diameter of 200 um, and achieve sensitivities of 1.83 uPa Hz^{-1/2}-10.4 mPa Hz^{-1/2} in 30 kHz-1.65 MHz range.

physics.optics↗

arXiv2022

Instance-wise Prompt Tuning for Pretrained Language Models

Prompt Learning has recently gained great popularity in bridging the gap between pretraining tasks and various downstream tasks. It freezes Pretrained Language Models (PLMs) and only tunes a few task-related parameters (prompts) for downstream tasks, greatly reducing the cost of tuning giant models. The key enabler of this is the idea of querying PLMs with task-specific knowledge implicated in prompts. This paper reveals a major limitation of existing methods that the indiscriminate prompts for all input data in a task ignore the intrinsic knowledge from input data, resulting in sub-optimal performance. We introduce Instance-wise Prompt Tuning (IPT), the first prompt learning paradigm that injects knowledge from the input data instances to the prompts, thereby providing PLMs with richer and more concrete context information. We devise a series of strategies to produce instance-wise prompts, addressing various concerns like model quality and cost-efficiency. Across multiple tasks and resource settings, IPT significantly outperforms task-based prompt learning methods, and achieves comparable performance to conventional finetuning with only 0.5% - 1.5% of tuned parameters.

cs.CL↗

arXiv2022

K-Detector: Identifying Duplicate Crash Failures in Large-Scale Software Delivery

After a developer submits code, corresponding test cases arise to ensure the quality of software delivery. Test failures would occur during this period, such as crash, error, and timeout. Since it takes time for developers to resolve them, many duplicate failures will happen during this period. In the delivery practice of SAP HANA, crash triage is considered as the most time-consuming task. If duplicate crash failures can be automatically identified, the degree of automation will be significantly enhanced. To find such duplicates, we propose a training-based mathematical model that utilizes component information of SAP HANA to achieve better crash similarity comparison. We implement our approach in a tool named Knowledge-based Detector (K-Detector), which is verified by 11,208 samples and performs 0.986 in AUC. Furthermore, we have deployed K-Detector to the production environment, and it can save 97% human efforts in crash triage as statistics.

cs.SE↗

arXiv2022

Gravitational-Wave Signatures of Chiral-Symmetric Technicolor

A chiral-symmetric technicolor model successfully reconciles the tension between electroweak precision tests and traditional technicolor models. Focusing on its simplest realization preserving the conventional Higgs mechanism, we study its primordial gravitational wave signatures originating from first order phase transitions in the early Universe. We found that abundant phase transition patterns arise from a physically viable parameter space. Besides, we have also found gravitational wave signals possibly visible by future experiments, such as LISA, BBO and u-DECIGO. Our results stress the importance of gravitational wave detectors in exploring new physics complementary to ground colliders in the multi-messenger astronomy era.

hep-ph↗

arXiv2022

Exploring Entity Interactions for Few-Shot Relation Learning (Student Abstract)

Few-shot relation learning refers to infer facts for relations with a limited number of observed triples. Existing metric-learning methods for this problem mostly neglect entity interactions within and between triples. In this paper, we explore this kind of fine-grained semantic meanings and propose our model TransAM. Specifically, we serialize reference entities and query entities into sequence and apply transformer structure with local-global attention to capture both intra- and inter-triple entity interactions. Experiments on two public benchmark datasets NELL-One and Wiki-One with 1-shot setting prove the effectiveness of TransAM.

cs.CL↗

arXiv2022

Neighbors Are Not Strangers: Improving Non-Autoregressive Translation under Low-Frequency Lexical Constraints

However, current autoregressive approaches suffer from high latency. In this paper, we focus on non-autoregressive translation (NAT) for this problem for its efficiency advantage. We identify that current constrained NAT models, which are based on iterative editing, do not handle low-frequency constraints well. To this end, we propose a plug-in algorithm for this line of work, i.e., Aligned Constrained Training (ACT), which alleviates this problem by familiarizing the model with the source-side context of the constraints. Experiments on the general and domain datasets show that our model improves over the backbone constrained NAT model in constraint preservation and translation quality, especially for rare constraints.

cs.CL↗

arXiv2022

Real-Time Neural Character Rendering with Pose-Guided Multiplane Images

We propose pose-guided multiplane image (MPI) synthesis which can render an animatable character in real scenes with photorealistic quality. We use a portable camera rig to capture the multi-view images along with the driving signal for the moving subject. Our method generalizes the image-to-image translation paradigm, which translates the human pose to a 3D scene representation -- MPIs that can be rendered in free viewpoints, using the multi-views captures as supervision. To fully cultivate the potential of MPI, we propose depth-adaptive MPI which can be learned using variable exposure images while being robust to inaccurate camera registration. Our method demonstrates advantageous novel-view synthesis quality over the state-of-the-art approaches for characters with challenging motions. Moreover, the proposed method is generalizable to novel combinations of training poses and can be explicitly controlled. Our method achieves such expressive and animatable character rendering all in real time, serving as a promising solution for practical applications.

cs.CV↗

arXiv2022

Large-Scale Pre-training for Person Re-identification with Noisy Labels

This paper aims to address the problem of pre-training for person re-identification (Re-ID) with noisy labels. To setup the pre-training task, we apply a simple online multi-object tracking system on raw videos of an existing unlabeled Re-ID dataset "LUPerson" nd build the Noisy Labeled variant called "LUPerson-NL". Since theses ID labels automatically derived from tracklets inevitably contain noises, we develop a large-scale Pre-training framework utilizing Noisy Labels (PNL), which consists of three learning modules: supervised Re-ID learning, prototype-based contrastive learning, and label-guided contrastive learning. In principle, joint learning of these three modules not only clusters similar examples to one prototype, but also rectifies noisy labels based on the prototype assignment. We demonstrate that learning directly from raw videos is a promising alternative for pre-training, which utilizes spatial and temporal correlations as weak supervision. This simple pre-training task provides a scalable way to learn SOTA Re-ID representations from scratch on "LUPerson-NL" without bells and whistles. For example, by applying on the same supervised Re-ID method MGN, our pre-trained model improves the mAP over the unsupervised pre-training counterpart by 5.7%, 2.2%, 2.3% on CUHK03, DukeMTMC, and MSMT17 respectively. Under the small-scale or few-shot setting, the performance gain is even more significant, suggesting a better transferability of the learned representation. Code is available at https://github.com/DengpanFu/LUPerson-NL

cs.CV↗