arXiv subjects

Zihao Wang

Publications and source records attributed to Zihao Wang.

At least 325 records · Page 18Linked to original sources

arXiv2022

Query2Particles: Knowledge Graph Reasoning with Particle Embeddings

Answering complex logical queries on incomplete knowledge graphs (KGs) with missing edges is a fundamental and important task for knowledge graph reasoning. The query embedding method is proposed to answer these queries by jointly encoding queries and entities to the same embedding space. Then the answer entities are selected according to the similarities between the entity embeddings and the query embedding. As the answers to a complex query are obtained from a combination of logical operations over sub-queries, the embeddings of the answer entities may not always follow a uni-modal distribution in the embedding space. Thus, it is challenging to simultaneously retrieve a set of diverse answers from the embedding space using a single and concentrated query representation such as a vector or a hyper-rectangle. To better cope with queries with diversified answers, we propose Query2Particles (Q2P), a complex KG query answering method. Q2P encodes each query into multiple vectors, named particle embeddings. By doing so, the candidate answers can be retrieved from different areas over the embedding space using the maximal similarities between the entity embeddings and any of the particle embeddings. Meanwhile, the corresponding neural logic operations are defined to support its reasoning over arbitrary first-order logic queries. The experiments show that Query2Particles achieves state-of-the-art performance on the complex query answering tasks on FB15k, FB15K-237, and NELL knowledge graphs.

cs.CL↗

arXiv2022

Utterance Rewriting with Contrastive Learning in Multi-turn Dialogue

Context modeling plays a significant role in building multi-turn dialogue systems. In order to make full use of context information, systems can use Incomplete Utterance Rewriting(IUR) methods to simplify the multi-turn dialogue into single-turn by merging current utterance and context information into a self-contained utterance. However, previous approaches ignore the intent consistency between the original query and rewritten query. The detection of omitted or coreferred locations in the original query can be further improved. In this paper, we introduce contrastive learning and multi-task learning to jointly model the problem. Our method benefits from carefully designed self-supervised objectives, which act as auxiliary tasks to capture semantics at both sentence-level and token-level. The experiments show that our proposed model achieves state-of-the-art performance on several public datasets.

cs.CL↗

arXiv2022

Machine Learning for Encrypted Malicious Traffic Detection: Approaches, Datasets and Comparative Study

As people's demand for personal privacy and data security becomes a priority, encrypted traffic has become mainstream in the cyber world. However, traffic encryption is also shielding malicious and illegal traffic introduced by adversaries, from being detected. This is especially so in the post-COVID-19 environment where malicious traffic encryption is growing rapidly. Common security solutions that rely on plain payload content analysis such as deep packet inspection are rendered useless. Thus, machine learning based approaches have become an important direction for encrypted malicious traffic detection. In this paper, we formulate a universal framework of machine learning based encrypted malicious traffic detection techniques and provided a systematic review. Furthermore, current research adopts different datasets to train their models due to the lack of well-recognized datasets and feature sets. As a result, their model performance cannot be compared and analyzed reliably. Therefore, in this paper, we analyse, process and combine datasets from 5 different sources to generate a comprehensive and fair dataset to aid future research in this field. On this basis, we also implement and compare 10 encrypted malicious traffic detection algorithms. We then discuss challenges and propose future directions of research.

cs.CR↗

arXiv2022

CLIP-GEN: Language-Free Training of a Text-to-Image Generator with CLIP

Training a text-to-image generator in the general domain (e.g., Dall.e, CogView) requires huge amounts of paired text-image data, which is too expensive to collect. In this paper, we propose a self-supervised scheme named as CLIP-GEN for general text-to-image generation with the language-image priors extracted with a pre-trained CLIP model. In our approach, we only require a set of unlabeled images in the general domain to train a text-to-image generator. Specifically, given an image without text labels, we first extract the embedding of the image in the united language-vision embedding space with the image encoder of CLIP. Next, we convert the image into a sequence of discrete tokens in the VQGAN codebook space (the VQGAN model can be trained with the unlabeled image dataset in hand). Finally, we train an autoregressive transformer that maps the image tokens from its unified language-vision representation. Once trained, the transformer can generate coherent image tokens based on the text embedding extracted from the text encoder of CLIP upon an input text. Such a strategy enables us to train a strong and general text-to-image generator with large text-free image dataset such as ImageNet. Qualitative and quantitative evaluations verify that our method significantly outperforms optimization-based text-to-image methods in terms of image quality while not compromising the text-image matching. Our method can even achieve comparable performance as flagship supervised models like CogView.

cs.CV↗

arXiv2022

Fast Sinkhorn I: An O(N) algorithm for the Wasserstein-1 metric

The Wasserstein metric is broadly used in optimal transport for comparing two probabilistic distributions, with successful applications in various fields such as machine learning, signal processing, seismic inversion, etc. Nevertheless, the high computational complexity is an obstacle for its practical applications. The Sinkhorn algorithm, one of the main methods in computing the Wasserstein metric, solves an entropy regularized minimizing problem, which allows arbitrary approximations to the Wasserstein metric with O(N^2) computational cost. However, higher accuracy of its numerical approximation requires more Sinkhorn iterations with repeated matrix-vector multiplications, which is still unaffordable. In this work, we propose an efficient implementation of the Sinkhorn algorithm to calculate the Wasserstein-1 metric with O(N) computational cost, which achieves the optimal theoretical complexity. By utilizing the special structure of Sinkhorn's kernel, the repeated matrix-vector multiplications can be implemented with O(N) times multiplications and additions, using the Qin Jiushao or Horner's method for efficient polynomial evaluation, leading to an efficient algorithm without losing accuracy. In addition, the log-domain stabilization technique, used to stabilize the iterative procedure, can also be applied in this algorithm. Our numerical experiments show that the newly developed algorithm is one to three orders of magnitude faster than the original Sinkhorn algorithm.

math.OC↗

arXiv2021

Mending Neural Implicit Modeling for 3D Vehicle Reconstruction in the Wild

Reconstructing high-quality 3D objects from sparse, partial observations from a single view is of crucial importance for various applications in computer vision, robotics, and graphics. While recent neural implicit modeling methods show promising results on synthetic or dense data, they perform poorly on sparse and noisy real-world data. We discover that the limitations of a popular neural implicit model are due to lack of robust shape priors and lack of proper regularization. In this work, we demonstrate highquality in-the-wild shape reconstruction using: (i) a deep encoder as a robust-initializer of the shape latent-code; (ii) regularized test-time optimization of the latent-code; (iii) a deep discriminator as a learned high-dimensional shape prior; (iv) a novel curriculum learning strategy that allows the model to learn shape priors on synthetic data and smoothly transfer them to sparse real world data. Our approach better captures the global structure, performs well on occluded and sparse observations, and registers well with the ground-truth shape. We demonstrate superior performance over state-of-the-art 3D object reconstruction methods on two real-world datasets.

cs.CV↗

arXiv2021

Benchmarking the Combinatorial Generalizability of Complex Query Answering on Knowledge Graphs

Complex Query Answering (CQA) is an important reasoning task on knowledge graphs. Current CQA learning models have been shown to be able to generalize from atomic operators to more complex formulas, which can be regarded as the combinatorial generalizability. In this paper, we present EFO-1-QA, a new dataset to benchmark the combinatorial generalizability of CQA models by including 301 different queries types, which is 20 times larger than existing datasets. Besides, our work, for the first time, provides a benchmark to evaluate and analyze the impact of different operators and normal forms by using (a) 7 choices of the operator systems and (b) 9 forms of complex queries. Specifically, we provide the detailed study of the combinatorial generalizability of two commonly used operators, i.e., projection and intersection, and justify the impact of the forms of queries given the canonical choice of operators. Our code and data can provide an effective pipeline to benchmark CQA models.

cs.CL↗

arXiv2021

Adversarial Attacks on ML Defense Models Competition

Due to the vulnerability of deep neural networks (DNNs) to adversarial examples, a large number of defense techniques have been proposed to alleviate this problem in recent years. However, the progress of building more robust models is usually hampered by the incomplete or incorrect robustness evaluation. To accelerate the research on reliable evaluation of adversarial robustness of the current defense models in image classification, the TSAIL group at Tsinghua University and the Alibaba Security group organized this competition along with a CVPR 2021 workshop on adversarial machine learning (https://aisecure-workshop.github.io/amlcvpr2021/). The purpose of this competition is to motivate novel attack algorithms to evaluate adversarial robustness more effectively and reliably. The participants were encouraged to develop stronger white-box attack algorithms to find the worst-case robustness of different defenses. This competition was conducted on an adversarial robustness evaluation platform -- ARES (https://github.com/thu-ml/ares), and is held on the TianChi platform (https://tianchi.aliyun.com/competition/entrance/531847/introduction) as one of the series of AI Security Challengers Program. After the competition, we summarized the results and established a new adversarial robustness benchmark at https://ml.cs.tsinghua.edu.cn/ares-bench/, which allows users to upload adversarial attack algorithms and defense models for evaluation.

cs.CV↗

arXiv2021

A Speaker-aware Parallel Hierarchical Attentive Encoder-Decoder Model for Multi-turn Dialogue Generation

This paper presents a novel open-domain dialogue generation model emphasizing the differentiation of speakers in multi-turn conversations. Differing from prior work that solely relies on the content of conversation history to generate a response, we argue that capturing relative social relations among utterances (i.e., generated by either the same speaker or different persons) benefits the machine capturing fine-grained context information from a conversation history to improve context coherence in the generated response. Given that, we propose a speaker-aware Parallel Hierarchical Attentive Encoder-Decoder (PHAED) model that aims to model each utterance with the awareness of its speaker and contextual associations with the same speaker's previous messages. Specifically, in a conversation involving two speakers, we regard the utterances from one speaker as responses and those from the other as queries. After understanding queries via our encoder with inner-query and inter-query encodings, our decoder reuses the hidden states of previously generated responses, instead of reconstructing these by the encoder, to generate a new response. Our empirical results show that PHAED outperforms the state-of-the-art in both automatic and human evaluations. Furthermore, our ablation study shows that dialogue models with speaker tokens can generally decrease the possibility of generating non-coherent responses regarding the conversation context.

cs.CL↗

arXiv2021

A Bombieri-Vinogradov theorem for higher rank groups

We establish a result of Bombieri-Vinogradov type for the Dirichlet coefficients at prime ideals of the standard $L$-function associated to a self-dual cuspidal automorphic representation $π$ of $\mathrm{GL}_n$ over a number field $F$ which is not a quadratic twist of itself. Our result does not rely on any unproven progress towards the generalized Ramanujan conjecture or the nonexistence of Landau-Siegel zeros. In particular, when $π$ is fixed and not equal to a quadratic twist of itself, we prove the first unconditional Siegel-type lower bound for the twisted $L$-values $|L(1,π\otimesχ)|$ in the $χ$-aspect, where $χ$ is a primitive quadratic Hecke character over $F$. Our result improves the levels of distribution in other works that relied on these unproven hypotheses. As applications, when $n=2,3,4$, we prove a $\mathrm{GL}_n$ analogue of the Titchmarsh divisor problem and a nontrivial bound for a certain $\mathrm{GL}_n\times\mathrm{GL}_2$ shifted convolution sum.

math.NT↗

arXiv2021

Projectively enriched symmetry and topology in acoustic crystals

Symmetry plays a key role in modern physics, as manifested in the revolutionary topological classification of matter in the past decade. So far, we seem to have a complete theory of topological phases from internal symmetries as well as crystallographic symmetry groups. However, an intrinsic element, i.e., the gauge symmetry in physical systems, has been overlooked in the current framework. Here, we show that the algebraic structure of crystal symmetries can be projectively enriched due to the gauge symmetry, which subsequently gives rise to new topological physics never witnessed under ordinary symmetries. We demonstrate the idea by theoretical analysis, numerical simulation, and experimental realization of a topological acoustic lattice with projective translation symmetries under a $Z_2$ gauge field, which exhibits unique features of rich topologies, including a single Dirac point, Möbius topological insulator and graphene-like semimetal phases on a rectangular lattice. Our work reveals the impact when gauge and crystal symmetries meet together with topology, and opens the door to a vast unexplored land of topological states by projective symmetries.

cond-mat.mes-hall↗

arXiv2021

Economic Recession Prediction Using Deep Neural Network

We investigate the effectiveness of different machine learning methodologies in predicting economic cycles. We identify the deep learning methodology of Bi-LSTM with Autoencoder as the most accurate model to forecast the beginning and end of economic recessions in the U.S. We adopt commonly-available macro and market-condition features to compare the ability of different machine learning models to generate good predictions both in-sample and out-of-sample. The proposed model is flexible and dynamic when both predictive variables and model coefficients vary over time. It provided good out-of-sample predictions for the past two recessions and early warning about the COVID-19 recession.

econ.GN↗

arXiv2021

Quasi-symplectic Langevin Variational Autoencoder

Variational autoencoder (VAE) is a very popular and well-investigated generative model in neural learning research. To leverage VAE in practical tasks dealing with a massive dataset of large dimensions, it is required to deal with the difficulty of building low variance evidence lower bounds (ELBO). Markov Chain Monte Carlo (MCMC) is an effective approach to tighten the ELBO for approximating the posterior distribution and Hamiltonian Variational Autoencoder (HVAE) is an effective MCMC inspired approach for constructing a low-variance ELBO that is amenable to the reparameterization trick. The HVAE adapted the Hamiltonian dynamic flow into variational inference that significantly improves the performance of the posterior estimation. We propose in this work a Langevin dynamic flow-based inference approach by incorporating the gradients information in the inference process through the Langevin dynamic which is a kind of MCMC based method similar to HVAE. Specifically, we employ a quasi-symplectic integrator to cope with the prohibit problem of the Hessian computing in naive Langevin flow. We show the theoretical and practical effectiveness of the proposed framework with other gradient flow-based methods.

stat.ML↗

arXiv2020

PV-NAS: Practical Neural Architecture Search for Video Recognition

Recently, deep learning has been utilized to solve video recognition problem due to its prominent representation ability. Deep neural networks for video tasks is highly customized and the design of such networks requires domain experts and costly trial and error tests. Recent advance in network architecture search has boosted the image recognition performance in a large margin. However, automatic designing of video recognition network is less explored. In this study, we propose a practical solution, namely Practical Video Neural Architecture Search (PV-NAS).Our PV-NAS can efficiently search across tremendous large scale of architectures in a novel spatial-temporal network search space using the gradient based search methods. To avoid sticking into sub-optimal solutions, we propose a novel learning rate scheduler to encourage sufficient network diversity of the searched models. Extensive empirical evaluations show that the proposed PV-NAS achieves state-of-the-art performance with much fewer computational resources. 1) Within light-weight models, our PV-NAS-L achieves 78.7% and 62.5% Top-1 accuracy on Kinetics-400 and Something-Something V2, which are better than previous state-of-the-art methods (i.e., TSM) with a large margin (4.6% and 3.4% on each dataset, respectively), and 2) among median-weight models, our PV-NAS-M achieves the best performance (also a new record)in the Something-Something V2 dataset.

cs.CV↗

arXiv2020

Long Short-Term Memory Neuron Equalizer

In this work we propose a neuromorphic hardware based signal equalizer by based on the deep learning implementation. The proposed neural equalizer is plasticity trainable equalizer which is different from traditional model designed based DFE. A trainable Long Short-Term memory neural network based DFE architecture is proposed for signal recovering and digital implementation is evaluated through FPGA implementation. Constructing with modelling based equalization methods, the proposed approach is compatible to multiple frequency signal equalization instead of single type signal equalization. We shows quantitatively that the neuronmorphic equalizer which is amenable both analog and digital implementation outperforms in different metrics in comparison with benchmarks approaches. The proposed method is adaptable both for general neuromorphic computing or ASIC instruments.

eess.SP↗

arXiv2020

Feature matching in Ultrasound images

Feature matching is an important technique to identify a single object in different images. It helps machines to construct recognition of a specific object from multiple perspectives. For years, feature matching has been commonly used in various computer vision applications, like traffic surveillance, self-driving, and other systems. With the arise of Computer-Aided Diagnosis(CAD), the need for feature matching techniques also emerges in the medical imaging field. In this paper, we present a deep learning-based method specially for ultrasound images. It will be examined against existing methods that have outstanding results on regular images. As the ultrasound images are different from regular images in many fields like texture, noise type, and dimension, traditional methods will be evaluated and optimized to be applied to ultrasound images.

cs.CV↗

arXiv2020

An Investigation of how Label Smoothing Affects Generalization

It has been hypothesized that label smoothing can reduce overfitting and improve generalization, and current empirical evidence seems to corroborate these effects. However, there is a lack of mathematical understanding of when and why such empirical improvements occur. In this paper, as a step towards understanding why label smoothing is effective, we propose a theoretical framework to show how label smoothing provides in controlling the generalization loss. In particular, we show that this benefit can be precisely formulated and identified in the label noise setting, where the training is partially mislabeled. Our theory also predicts the existence of an optimal label smoothing point, a single value for the label smoothing hyperparameter that minimizes generalization loss. Extensive experiments are done to confirm the predictions of our theory. We believe that our findings will help both theoreticians and practitioners understand label smoothing, and better apply them to real-world datasets.

cs.LG↗

arXiv2020

A Relaxed Matching Procedure for Unsupervised BLI

Recently unsupervised Bilingual Lexicon Induction (BLI) without any parallel corpus has attracted much research interest. One of the crucial parts in methods for the BLI task is the matching procedure. Previous works impose a too strong constraint on the matching and lead to many counterintuitive translation pairings. Thus, We propose a relaxed matching procedure to find a more precise matching between two languages. We also find that aligning source and target language embedding space bidirectionally will bring significant improvement. We follow the previous iterative framework to conduct experiments. Results on standard benchmark demonstrate the effectiveness of our proposed method, which substantially outperforms previous unsupervised methods.

cs.CL↗