arXiv subjects

Hao Tian

Publications and source records attributed to Hao Tian.

At least 127 records · Page 7Linked to original sources

arXiv2023

A fast computational framework for the linear bond-based peridynamic model

Peridynamic (PD) theory is significant and promising in engineering and materials science; however, it imposes challenges owing to the enormous computational cost caused by its nonlocality. Our main contribution, which overcomes the restrictions of the existing fast method, is a general computational framework for the linear bond-based peridynamic models based on the meshfree method, called the matrix-structure-based fast method (MSBFM), which is suitable for the general case, including 2D/3D problems, and static/dynamic issues, as well as problems with general boundary conditions, in particular, problems with crack propagation. Consequently, we provide a general calculation flow chart. The proposed computational framework is practical and easily embedded into the existing computational algorithm. With this framework, the computational cost is reduced from $O(N^2)$ to $O(N\log N)$, and the storage request is reduced from $O(N^2)$ to $O(N)$, where N is the degree of freedom. Finally, the vast reduction of the computational and memory requirement is verified by numerical examples.

math.NA↗

arXiv2023

ERNIE 3.0 Tiny: Frustratingly Simple Method to Improve Task-Agnostic Distillation Generalization

Task-agnostic knowledge distillation attempts to address the problem of deploying large pretrained language model in resource-constrained scenarios by compressing a large pretrained model called teacher into a smaller one called student such that the student can be directly finetuned on downstream tasks and retains comparable performance. However, we empirically find that there is a generalization gap between the student and the teacher in existing methods. In this work, we show that we can leverage multi-task learning in task-agnostic distillation to advance the generalization of the resulted student. In particular, we propose Multi-task Infused Task-agnostic Knowledge Distillation (MITKD). We first enhance the teacher by multi-task training it on multiple downstream tasks and then perform distillation to produce the student. Experimental results demonstrate that our method yields a student with much better generalization, significantly outperforms existing baselines, and establishes a new state-of-the-art result on in-domain, out-domain, and low-resource datasets in the setting of task-agnostic distillation. Moreover, our method even exceeds an 8x larger BERT$_{\text{Base}}$ on SQuAD and four GLUE tasks. In addition, by combining ERNIE 3.0, our method achieves state-of-the-art results on 10 Chinese datasets.

cs.CL↗

arXiv2023

An Integrated Photon-Pair Source with Monolithic Piezoelectric Frequency Tunability

This work demonstrates the capabilities of an entangled photon-pair source at telecom wavelengths, based on a photonic integrated Si$_3$N$_4$ microresonator with monolithically integrated piezoelectric frequency tuning. Previously, frequency tuning of photon-pairs generated by microresonators has only been demonstrated using thermal control, however these have limited actuation bandwidth, and are not compatible with cryogenic environments. Here, the frequency-tunable photon-pair generation capabilities of a Si$_3$N$_4$ microresonator with a monolithically integrated aluminium nitride layer are shown. Fast-frequency locking of the microresonator to an external laser is demonstrated, with a resulting locking bandwidth orders of magnitude larger than reported previously using thermal locking. These abilities will have direct application in future schemes which interface such sources with quantum memories based on e.g. trapped-ion or rare-earth ion schemes.

quant-ph↗

arXiv2022

LAMOST medium-resolution spectroscopic survey of binarity and exotic star (LAMOST-MRS-B): Observation strategy and target selection

LAMOST-MRS-B is one of the sub-surveys of LAMOST medium-resolution (R~7500) spectroscopic survey. It aims at studying the statistical properties (e.g., binary fraction, orbital period distribution, mass ratio distribution) of binary stars and exotic stars. We intend to observe about 30000 stars (10 mag <= G <= 14.5 mag) with at least 10 visits in five years. We first planned to observe 25 plates around the galactic plane in 2018. Then the plates were reduced to 12 in 2019 because of the limitation of observation. At the same time, two new plates located at the high galactic latitude were added to explore binary properties influenced by the different environments. In this survey project, we set the identified exotic and low-metallicity stars with the highest observation priorities. For the rest of the selected stars, we gave higher priority to the relatively brighter stars in order to obtain high-quality spectra as many as possible. Spectra of 49129 stars have been obtained in LAMOST-MRS-B field and released in DR8, of which 28828 and 3375 stars have been visited more than twice and ten times with SNR >= 10, respectively. Most of the sources are B-, A-, and F-type stars with 0.6 < [Fe/H] < 0.4 dex. We also obtain 347 identified variable and exotic stars and about 250 stars with [Fe/H] < 1 dex. We measure radial velocities (RVs) by using 892233 spectra of the stars. The uncertainties of RV achieve about 1 km/s and 10 km/s1 for 95% of late- and early-type stars, respectively. The datasets presented in this paper are available at http://www.doi.org/10.57760/sciencedb.j00113.00035.

astro-ph.SR↗

arXiv2022

BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View Recognition via Perspective Supervision

We present a novel bird's-eye-view (BEV) detector with perspective supervision, which converges faster and better suits modern image backbones. Existing state-of-the-art BEV detectors are often tied to certain depth pre-trained backbones like VoVNet, hindering the synergy between booming image backbones and BEV detectors. To address this limitation, we prioritize easing the optimization of BEV detectors by introducing perspective space supervision. To this end, we propose a two-stage BEV detector, where proposals from the perspective head are fed into the bird's-eye-view head for final predictions. To evaluate the effectiveness of our model, we conduct extensive ablation studies focusing on the form of supervision and the generality of the proposed detector. The proposed method is verified with a wide spectrum of traditional and modern image backbones and achieves new SoTA results on the large-scale nuScenes dataset. The code shall be released soon.

cs.CV↗

arXiv2022

Is the core-cusp problem a matter of perspective: Jeans Anisotropic Modeling against numerical simulations

Mock member stars for 28 dwarf galaxies are constructed from the cosmological Auriga simulation, which reflect the dynamical status of realistic stellar tracers. The axis-symmetric Jeans Anisotropic Multi-Gaussian Expansion (JAM) modeling is applied to 6,000 star particles for each system, to recover the underlying matter distribution. The stellar or dark matter component individually is poorly recovered, but the total profile is constrained more reasonably. The mass within the half-mass radius of tracers is recovered the tightest, and the mass between 200 and 300 pc, $M(200-300\mathrm{pc})$, is constrained ensemble unbiasedly, with a scatter of 0.167 dex. If using 2,000 particles and only line-of-sight velocities with typical errors, the scatter in $M(200-300\mathrm{pc})$ is increased by $\sim$50%. Quiescent Sagittarius dSph-like systems and star-forming systems with strong outflows show distinct features, with $M(200-300\mathrm{pc})$ mostly under-estimated for the former, and likely over-estimated for the latter. The biases correlate with the dynamical status, which is a result of contraction motions due to tidal effects in quiescent systems or galactic winds in star-forming systems, driving them out of equilibrium. After including Gaia DR3 proper motion errors, we find proper motions can be as useful as line-of-sight velocities for nearby systems at $<\sim$60 kpc. By extrapolating the actual density profiles and the dynamical constraints down to scales below the resolution, we find the mass within 150 pc can be constrained ensemble unbiasedly, with a scatter of $\sim$0.255 dex. In the end, we show that the contraction of member stars in nearby systems is detectable based on Gaia DR3 proper motion errors.

astro-ph.GA↗

arXiv2022

Arbitrary Style Guidance for Enhanced Diffusion-Based Text-to-Image Generation

Diffusion-based text-to-image generation models like GLIDE and DALLE-2 have gained wide success recently for their superior performance in turning complex text inputs into images of high quality and wide diversity. In particular, they are proven to be very powerful in creating graphic arts of various formats and styles. Although current models supported specifying style formats like oil painting or pencil drawing, fine-grained style features like color distributions and brush strokes are hard to specify as they are randomly picked from a conditional distribution based on the given text input. Here we propose a novel style guidance method to support generating images using arbitrary style guided by a reference image. The generation method does not require a separate style transfer model to generate desired styles while maintaining image quality in generated content as controlled by the text input. Additionally, the guidance method can be applied without a style reference, denoted as self style guidance, to generate images of more diverse styles. Comprehensive experiments prove that the proposed method remains robust and effective in a wide range of conditions, including diverse graphic art forms, image content types and diffusion models.

cs.CV↗

arXiv2022

Extreme Generative Image Compression by Learning Text Embedding from Diffusion Models

Transferring large amount of high resolution images over limited bandwidth is an important but very challenging task. Compressing images using extremely low bitrates (<0.1 bpp) has been studied but it often results in low quality images of heavy artifacts due to the strong constraint in the number of bits available for the compressed data. It is often said that a picture is worth a thousand words but on the other hand, language is very powerful in capturing the essence of an image using short descriptions. With the recent success of diffusion models for text-to-image generation, we propose a generative image compression method that demonstrates the potential of saving an image as a short text embedding which in turn can be used to generate high-fidelity images which is equivalent to the original one perceptually. For a given image, its corresponding text embedding is learned using the same optimization process as the text-to-image diffusion model itself, using a learnable text embedding as input after bypassing the original transformer. The optimization is applied together with a learning compression model to achieve extreme compression of low bitrates <0.1 bpp. Based on our experiments measured by a comprehensive set of image quality metrics, our method outperforms the other state-of-the-art deep learning methods in terms of both perceptual quality and diversity.

eess.IV↗

arXiv2022

ERNIE-UniX2: A Unified Cross-lingual Cross-modal Framework for Understanding and Generation

Recent cross-lingual cross-modal works attempt to extend Vision-Language Pre-training (VLP) models to non-English inputs and achieve impressive performance. However, these models focus only on understanding tasks utilizing encoder-only architecture. In this paper, we propose ERNIE-UniX2, a unified cross-lingual cross-modal pre-training framework for both generation and understanding tasks. ERNIE-UniX2 integrates multiple pre-training paradigms (e.g., contrastive learning and language modeling) based on encoder-decoder architecture and attempts to learn a better joint representation across languages and modalities. Furthermore, ERNIE-UniX2 can be seamlessly fine-tuned for varieties of generation and understanding downstream tasks. Pre-trained on both multilingual text-only and image-text datasets, ERNIE-UniX2 achieves SOTA results on various cross-lingual cross-modal generation and understanding tasks such as multimodal machine translation and multilingual visual question answering.

cs.CV↗

arXiv2022

Clip-Tuning: Towards Derivative-free Prompt Learning with a Mixture of Rewards

Derivative-free prompt learning has emerged as a lightweight alternative to prompt tuning, which only requires model inference to optimize the prompts. However, existing work did not take full advantage of the over-parameterized characteristics of large pre-trained language models (PLMs). In this paper, we propose Clip-Tuning, a simple yet effective method that adopts diverse frozen "thinned" networks of PLMs to obtain a mixture of rewards and thus advance the derivative-free prompt learning. The thinned networks consist of all the hidden units that survive a stationary dropout strategy, whose inference predictions reflect an ensemble of partial views over prompted training samples. Our method outperforms previous gradient-free prompt learning methods and achieves parity with gradient-based counterparts on seven language understanding benchmarks under few-shot settings.

cs.CL↗

arXiv2022

ERNIE-Layout: Layout Knowledge Enhanced Pre-training for Visually-rich Document Understanding

Recent years have witnessed the rise and success of pre-training techniques in visually-rich document understanding. However, most existing methods lack the systematic mining and utilization of layout-centered knowledge, leading to sub-optimal performances. In this paper, we propose ERNIE-Layout, a novel document pre-training solution with layout knowledge enhancement in the whole workflow, to learn better representations that combine the features from text, layout, and image. Specifically, we first rearrange input sequences in the serialization stage, and then present a correlative pre-training task, reading order prediction, to learn the proper reading order of documents. To improve the layout awareness of the model, we integrate a spatial-aware disentangled attention into the multi-modal transformer and a replaced regions prediction task into the pre-training phase. Experimental results show that ERNIE-Layout achieves superior performance on various downstream tasks, setting new state-of-the-art on key information extraction, document image classification, and document question answering datasets. The code and models are publicly available at http://github.com/PaddlePaddle/PaddleNLP/tree/develop/model_zoo/ernie-layout.

cs.CL↗

arXiv2022

ERNIE-ViL 2.0: Multi-view Contrastive Learning for Image-Text Pre-training

Recent Vision-Language Pre-trained (VLP) models based on dual encoder have attracted extensive attention from academia and industry due to their superior performance on various cross-modal tasks and high computational efficiency. They attempt to learn cross-modal representation using contrastive learning on image-text pairs, however, the built inter-modal correlations only rely on a single view for each modality. Actually, an image or a text contains various potential views, just as humans could capture a real-world scene via diverse descriptions or photos. In this paper, we propose ERNIE-ViL 2.0, a Multi-View Contrastive learning framework to build intra-modal and inter-modal correlations between diverse views simultaneously, aiming at learning a more robust cross-modal representation. Specifically, we construct multiple views within each modality to learn the intra-modal correlation for enhancing the single-modal representation. Besides the inherent visual/textual views, we construct sequences of object tags as a special textual view to narrow the cross-modal semantic gap on noisy image-text pairs. Pre-trained with 29M publicly available datasets, ERNIE-ViL 2.0 achieves competitive results on English cross-modal retrieval. Additionally, to generalize our method to Chinese cross-modal tasks, we train ERNIE-ViL 2.0 through scaling up the pre-training datasets to 1.5B Chinese image-text pairs, resulting in significant improvements compared to previous SOTA results on Chinese cross-modal retrieval. We release our pre-trained models in https://github.com/PaddlePaddle/ERNIE.

cs.CV↗

arXiv2022

Evolving Decomposed Plasticity Rules for Information-Bottlenecked Meta-Learning

Artificial neural networks (ANNs) are typically confined to accomplishing pre-defined tasks by learning a set of static parameters. In contrast, biological neural networks (BNNs) can adapt to various new tasks by continually updating the neural connections based on the inputs, which is aligned with the paradigm of learning effective learning rules in addition to static parameters, e.g., meta-learning. Among various biologically inspired learning rules, Hebbian plasticity updates the neural network weights using local signals without the guide of an explicit target function, thus enabling an agent to learn automatically without human efforts. However, typical plastic ANNs using a large amount of meta-parameters violate the nature of the genomics bottleneck and potentially deteriorate the generalization capacity. This work proposes a new learning paradigm decomposing those connection-dependent plasticity rules into neuron-dependent rules thus accommodating $Θ(n^2)$ learnable parameters with only $Θ(n)$ meta-parameters. We also thoroughly study the effect of different neural modulation on plasticity. Our algorithms are tested in challenging random 2D maze environments, where the agents have to use their past experiences to shape the neural connections and improve their performances for the future. The results of our experiment validate the following: 1. Plasticity can be adopted to continually update a randomly initialized RNN to surpass pre-trained, more sophisticated recurrent models, especially when coming to long-term memorization. 2. Following the genomics bottleneck, the proposed decomposed plasticity can be comparable to or even more effective than canonical plasticity rules in some instances.

cs.AI↗

arXiv2022

Accurate ADMET Prediction with XGBoost

The absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties are important in drug discovery as they define efficacy and safety. In this work, we applied an ensemble of features, including fingerprints and descriptors, and a tree-based machine learning model, extreme gradient boosting, for accurate ADMET prediction. Our model performs well in the Therapeutics Data Commons ADMET benchmark group. For 22 tasks, our model is ranked first in 18 tasks and top 3 in 21 tasks. The trained machine learning models are integrated in ADMETboost, a web server that is publicly available at https://ai-druglab.smu.edu/admet.

q-bio.BM↗

arXiv2022

ERNIE-Search: Bridging Cross-Encoder with Dual-Encoder via Self On-the-fly Distillation for Dense Passage Retrieval

Neural retrievers based on pre-trained language models (PLMs), such as dual-encoders, have achieved promising performance on the task of open-domain question answering (QA). Their effectiveness can further reach new state-of-the-arts by incorporating cross-architecture knowledge distillation. However, most of the existing studies just directly apply conventional distillation methods. They fail to consider the particular situation where the teacher and student have different structures. In this paper, we propose a novel distillation method that significantly advances cross-architecture distillation for dual-encoders. Our method 1) introduces a self on-the-fly distillation method that can effectively distill late interaction (i.e., ColBERT) to vanilla dual-encoder, and 2) incorporates a cascade distillation process to further improve the performance with a cross-encoder teacher. Extensive experiments are conducted to validate that our proposed solution outperforms strong baselines and establish a new state-of-the-art on open-domain QA benchmarks.

cs.CL↗

arXiv2022

LAST: Latent Space Assisted Adaptive Sampling for Protein Trajectories

Molecular dynamics (MD) simulation is widely used to study protein conformations and dynamics. However, conventional simulation suffers from being trapped in some local energy minima that are hard to escape. Thus, most computational time is spent sampling in the already visited regions. This leads to an inefficient sampling process and further hinders the exploration of protein movements in affordable simulation time. The advancement of deep learning provides new opportunities for protein sampling. Variational autoencoders are a class of deep learning models to learn a low-dimensional representation (referred to as the latent space) that can capture the key features of the input data. Based on this characteristic, we proposed a new adaptive sampling method, latent space assisted adaptive sampling for protein trajectories (LAST), to accelerate the exploration of protein conformational space. This method comprises cycles of (i) variational autoencoders training, (ii) seed structure selection on the latent space and (iii) conformational sampling through additional MD simulations. The proposed approach is validated through the sampling of four structures of two protein systems: two metastable states of E. Coli adenosine kinase (ADK) and two native states of Vivid (VVD). In all four conformations, seed structures were shown to lie on the boundary of conformation distributions. Moreover, large conformational changes were observed in a shorter simulation time when compared with conventional MD (cMD) simulations in both systems. In metastable ADK simulations, LAST explored two transition paths toward two stable states while cMD became trapped in an energy basin. In VVD light state simulations, LAST was three times faster than cMD simulation with a similar conformational space.

q-bio.QM↗

arXiv2022

Searching Extra-tidal Features around the Globular Cluster Whiting 1

Whiting 1 is a faint and young globular cluster in the halo of the Milky Way, and was suggested to have originated in the Sagittarius spherical dwarf galaxy (Sgr dSph). In this paper, we use the deep DESI Legacy Imaging Surveys to explore tentative spatial connection between Whiting 1 and the Sgr dSph. We redetermine the fundamental parameters of Whiting 1 and use the best-fitting isochrone (age $τ$=6.5 Gyr, metalicity Z=0.005 and $\rm d_{\odot}$=26.9 kpc) to construct a theoretical matched filter for the extra-tidal features searching. Without any smooth technique to the matched filter density map, we detect a round-shape feature with possible leading and trailing tails on either side of the cluster. This raw image is not totally new compared to old discoveries, but confirms that no more large-scale features can be detected under a depth of r<=22.5 mag. In our results, the whole feature stretches 0.1-0.2 degree along the orbit of Whiting 1, which gives a much larger area than the cluster core. The tails on both sides of the cluster align along the orbital direction of the Sgr dSph as well as the cluster itself, which implies that these debris are probably stripped remnants of Whiting 1 by the Milky Way.

astro-ph.GA↗

arXiv2022

ERNIE-SPARSE: Learning Hierarchical Efficient Transformer Through Regularized Self-Attention

Sparse Transformer has recently attracted a lot of attention since the ability for reducing the quadratic dependency on the sequence length. We argue that two factors, information bottleneck sensitivity and inconsistency between different attention topologies, could affect the performance of the Sparse Transformer. This paper proposes a well-designed model named ERNIE-Sparse. It consists of two distinctive parts: (i) Hierarchical Sparse Transformer (HST) to sequentially unify local and global information. (ii) Self-Attention Regularization (SAR) method, a novel regularization designed to minimize the distance for transformers with different attention topologies. To evaluate the effectiveness of ERNIE-Sparse, we perform extensive evaluations. Firstly, we perform experiments on a multi-modal long sequence modeling task benchmark, Long Range Arena (LRA). Experimental results demonstrate that ERNIE-Sparse significantly outperforms a variety of strong baseline methods including the dense attention and other efficient sparse attention methods and achieves improvements by 2.77% (57.78% vs. 55.01%). Secondly, to further show the effectiveness of our method, we pretrain ERNIE-Sparse and verified it on 3 text classification and 2 QA downstream tasks, achieve improvements on classification benchmark by 0.83% (92.46% vs. 91.63%), on QA benchmark by 3.24% (74.67% vs. 71.43%). Experimental results continue to demonstrate its superior performance.

cs.CL↗