arXiv subjects

Pengfei Hu

Publications and source records attributed to Pengfei Hu.

At least 73 records · Page 4Linked to original sources

arXiv2023

HRDoc: Dataset and Baseline Method Toward Hierarchical Reconstruction of Document Structures

The problem of document structure reconstruction refers to converting digital or scanned documents into corresponding semantic structures. Most existing works mainly focus on splitting the boundary of each element in a single document page, neglecting the reconstruction of semantic structure in multi-page documents. This paper introduces hierarchical reconstruction of document structures as a novel task suitable for NLP and CV fields. To better evaluate the system performance on the new task, we built a large-scale dataset named HRDoc, which consists of 2,500 multi-page documents with nearly 2 million semantic units. Every document in HRDoc has line-level annotations including categories and relations obtained from rule-based extractors and human annotators. Moreover, we proposed an encoder-decoder-based hierarchical document structure parsing system (DSPS) to tackle this problem. By adopting a multi-modal bidirectional encoder and a structure-aware GRU decoder with soft-mask operation, the DSPS model surpass the baseline method by a large margin. All scripts and datasets will be made publicly available at https://github.com/jfma-USTC/HRDoc.

cs.CL↗

arXiv2022

Multimodal Tree Decoder for Table of Contents Extraction in Document Images

Table of contents (ToC) extraction aims to extract headings of different levels in documents to better understand the outline of the contents, which can be widely used for document understanding and information retrieval. Existing works often use hand-crafted features and predefined rule-based functions to detect headings and resolve the hierarchical relationship between headings. Both the benchmark and research based on deep learning are still limited. Accordingly, in this paper, we first introduce a standard dataset, HierDoc, including image samples from 650 documents of scientific papers with their content labels. Then we propose a novel end-to-end model by using the multimodal tree decoder (MTD) for ToC as a benchmark for HierDoc. The MTD model is mainly composed of three parts, namely encoder, classifier, and decoder. The encoder fuses the multimodality features of vision, text, and layout information for each entity of the document. Then the classifier recognizes and selects the heading entities. Next, to parse the hierarchical relationship between the heading entities, a tree-structured decoder is designed. To evaluate the performance, both the metric of tree-edit-distance similarity (TEDS) and F1-Measure are adopted. Finally, our MTD approach achieves an average TEDS of 87.2% and an average F1-Measure of 88.1% on the test set of HierDoc. The code and dataset will be released at: https://github.com/Pengfei-Hu/MTD.

cs.CV↗

arXiv2022

AccEar: Accelerometer Acoustic Eavesdropping with Unconstrained Vocabulary

With the increasing popularity of voice-based applications, acoustic eavesdropping has become a serious threat to users' privacy. While on smartphones the access to microphones needs an explicit user permission, acoustic eavesdropping attacks can rely on motion sensors (such as accelerometer and gyroscope), which access is unrestricted. However, previous instances of such attacks can only recognize a limited set of pre-trained words or phrases. In this paper, we present AccEar, an accelerometerbased acoustic eavesdropping attack that can reconstruct any audio played on the smartphone's loudspeaker with unconstrained vocabulary. We show that an attacker can employ a conditional Generative Adversarial Network (cGAN) to reconstruct highfidelity audio from low-frequency accelerometer signals. The presented cGAN model learns to recreate high-frequency components of the user's voice from low-frequency accelerometer signals through spectrogram enhancement. We assess the feasibility and effectiveness of AccEar attack in a thorough set of experiments using audio from 16 public personalities. As shown by the results in both objective and subjective evaluations, AccEar successfully reconstructs user speeches from accelerometer signals in different scenarios including varying sampling rate, audio volume, device model, etc.

cs.SD↗

arXiv2022

Federated Learning Hyper-Parameter Tuning from a System Perspective

Federated learning (FL) is a distributed model training paradigm that preserves clients' data privacy. It has gained tremendous attention from both academia and industry. FL hyper-parameters (e.g., the number of selected clients and the number of training passes) significantly affect the training overhead in terms of computation time, transmission time, computation load, and transmission load. However, the current practice of manually selecting FL hyper-parameters imposes a heavy burden on FL practitioners because applications have different training preferences. In this paper, we propose FedTune, an automatic FL hyper-parameter tuning algorithm tailored to applications' diverse system requirements in FL training. FedTune iteratively adjusts FL hyper-parameters during FL training and can be easily integrated into existing FL systems. Through extensive evaluations of FedTune for diverse applications and FL aggregation algorithms, we show that FedTune is lightweight and effective, achieving 8.48%-26.75% system overhead reduction compared to using fixed FL hyper-parameters. This paper assists FL practitioners in designing high-performance FL training solutions. The source code of FedTune is available at https://github.com/DataSysTech/FedTune.

cs.LG↗

arXiv2022

MFA-Conformer: Multi-scale Feature Aggregation Conformer for Automatic Speaker Verification

In this paper, we present Multi-scale Feature Aggregation Conformer (MFA-Conformer), an easy-to-implement, simple but effective backbone for automatic speaker verification based on the Convolution-augmented Transformer (Conformer). The architecture of the MFA-Conformer is inspired by recent stateof-the-art models in speech recognition and speaker verification. Firstly, we introduce a convolution subsampling layer to decrease the computational cost of the model. Secondly, we adopt Conformer blocks which combine Transformers and convolution neural networks (CNNs) to capture global and local features effectively. Finally, the output feature maps from all Conformer blocks are concatenated to aggregate multi-scale representations before final pooling. We evaluate the MFA-Conformer on the widely used benchmarks. The best system obtains 0.64%, 1.29% and 1.63% EER on VoxCeleb1-O, SITW.Dev, and SITW.Eval set, respectively. MFA-Conformer significantly outperforms the popular ECAPA-TDNN systems in both recognition performance and inference speed. Last but not the least, the ablation studies clearly demonstrate that the combination of global and local feature learning can lead to robust and accurate speaker embedding extraction. We have also released the code for future comparison.

cs.SD↗

arXiv2022

Learning Audio-Visual embedding for Person Verification in the Wild

It has already been observed that audio-visual embedding is more robust than uni-modality embedding for person verification. Here, we proposed a novel audio-visual strategy that considers aggregators from a fusion perspective. First, we introduced weight-enhanced attentive statistics pooling for the first time in face verification. We find that a strong correlation exists between modalities during pooling, so joint attentive pooling is proposed which contains cycle consistency to learn the implicit inter-frame weight. Finally, each modality is fused with a gated attention mechanism to gain robust audio-visual embedding. All the proposed models are trained on the VoxCeleb2 dev dataset and the best system obtains 0.18%, 0.27%, and 0.49% EER on three official trial lists of VoxCeleb1 respectively, which is to our knowledge the best-published results for person verification.

cs.CV↗

arXiv2022

High Speed Rotation Estimation with Dynamic Vision Sensors

Rotational speed is one of the important metrics to be measured for calibrating the electric motors in manufacturing, monitoring engine during car repairing, faults detection on electrical appliance and etc. However, existing measurement techniques either require prohibitive hardware (e.g., high-speed camera) or are inconvenient to use in real-world application scenarios. In this paper, we propose, EV-Tach, an event-based tachometer via efficient dynamic vision sensing on mobile devices. EV-Tach is designed as a high-fidelity and convenient tachometer by introducing dynamic vision sensor as a new sensing modality to capture the high-speed rotation precisely under various real-world scenarios. By designing a series of signal processing algorithms bespoke for dynamic vision sensing on mobile devices, EV-Tach is able to extract the rotational speed accurately from the event stream produced by dynamic vision sensing on rotary targets. According to our extensive evaluations, the Relative Mean Absolute Error (RMAE) of EV-Tach is as low as 0.03% which is comparable to the state-of-the-art laser tachometer under fixed measurement mode. Moreover, EV-Tach is robust to subtle movement of user's hand, therefore, can be used as a handheld device, where the laser tachometer fails to produce reasonable results.

cs.CV↗

arXiv2022

PM-MMUT: Boosted Phone-Mask Data Augmentation using Multi-Modeling Unit Training for Phonetic-Reduction-Robust E2E Speech Recognition

Consonant and vowel reduction are often encountered in speech, which might cause performance degradation in automatic speech recognition (ASR). Our recently proposed learning strategy based on masking, Phone Masking Training (PMT), alleviates the impact of such phenomenon in Uyghur ASR. Although PMT achieves remarkably improvements, there still exists room for further gains due to the granularity mismatch between the masking unit of PMT (phoneme) and the modeling unit (word-piece). To boost the performance of PMT, we propose multi-modeling unit training (MMUT) architecture fusion with PMT (PM-MMUT). The idea of MMUT framework is to split the Encoder into two parts including acoustic feature sequences to phoneme-level representation (AF-to-PLR) and phoneme-level representation to word-piece-level representation (PLR-to-WPLR). It allows AF-to-PLR to be optimized by an intermediate phoneme-based CTC loss to learn the rich phoneme-level context information brought by PMT. Experimental results on Uyghur ASR show that the proposed approaches outperform obviously the pure PMT. We also conduct experiments on the 960-hour Librispeech benchmark using ESPnet1, which achieves about 10% relative WER reduction on all the test set without LM fusion comparing with the latest official ESPnet1 pre-trained model.

cs.SD↗

arXiv2022

Linguistic-Acoustic Similarity Based Accent Shift for Accent Recognition

General accent recognition (AR) models tend to directly extract low-level information from spectrums, which always significantly overfit on speakers or channels. Considering accent can be regarded as a series of shifts relative to native pronunciation, distinguishing accents will be an easier task with accent shift as input. But due to the lack of native utterance as an anchor, estimating the accent shift is difficult. In this paper, we propose linguistic-acoustic similarity based accent shift (LASAS) for AR tasks. For an accent speech utterance, after mapping the corresponding text vector to multiple accent-associated spaces as anchors, its accent shift could be estimated by the similarities between the acoustic embedding and those anchors. Then, we concatenate the accent shift with a dimension-reduced text vector to obtain a linguistic-acoustic bimodal representation. Compared with pure acoustic embedding, the bimodal representation is richer and more clear by taking full advantage of both linguistic and acoustic information, which can effectively improve AR performance. Experiments on Accented English Speech Recognition Challenge (AESRC) dataset show that our method achieves 77.42% accuracy on Test set, obtaining a 6.94% relative improvement over a competitive system in the challenge.

cs.SD↗

arXiv2022

Defensive Patches for Robust Recognition in the Physical World

To operate in real-world high-stakes environments, deep learning systems have to endure noises that have been continuously thwarting their robustness. Data-end defense, which improves robustness by operations on input data instead of modifying models, has attracted intensive attention due to its feasibility in practice. However, previous data-end defenses show low generalization against diverse noises and weak transferability across multiple models. Motivated by the fact that robust recognition depends on both local and global features, we propose a defensive patch generation framework to address these problems by helping models better exploit these features. For the generalization against diverse noises, we inject class-specific identifiable patterns into a confined local patch prior, so that defensive patches could preserve more recognizable features towards specific classes, leading models for better recognition under noises. For the transferability across multiple models, we guide the defensive patches to capture more global feature correlations within a class, so that they could activate model-shared global perceptions and transfer better among models. Our defensive patches show great potentials to improve application robustness in practice by simply sticking them around target objects. Extensive experiments show that we outperform others by large margins (improve 20+\% accuracy for both adversarial and corruption robustness on average in the digital and physical world). Our codes are available at https://github.com/nlsde-safety-team/DefensivePatch

cs.CV↗

arXiv2022

Leveraging Phone Mask Training for Phonetic-Reduction-Robust E2E Uyghur Speech Recognition

In Uyghur speech, consonant and vowel reduction are often encountered, especially in spontaneous speech with high speech rate, which will cause a degradation of speech recognition performance. To solve this problem, we propose an effective phone mask training method for Conformer-based Uyghur end-to-end (E2E) speech recognition. The idea is to randomly mask off a certain percentage features of phones during model training, which simulates the above verbal phenomena and facilitates E2E model to learn more contextual information. According to experiments, the above issues can be greatly alleviated. In addition, deep investigations are carried out into different units in masking, which shows the effectiveness of our proposed masking unit. We also further study the masking method and optimize filling strategy of phone mask. Finally, compared with Conformer-based E2E baseline without mask training, our model demonstrates about 5.51% relative Word Error Rate (WER) reduction on reading speech and 12.92% on spontaneous speech, respectively. The above approach has also been verified on test-set of open-source data THUYG-20, which shows 20% relative improvements.

cs.SD↗

arXiv2022

Iota: A Framework for Analyzing System-Level Security of IoTs

Most IoT systems involve IoT devices, communication protocols, remote cloud, IoT applications, mobile apps, and the physical environment. However, existing IoT security analyses only focus on a subset of all the essential components, such as device firmware, and ignore IoT systems' interactive nature, resulting in limited attack detection capabilities. In this work, we propose Iota, a logic programming-based framework to perform system-level security analysis for IoT systems. Iota generates attack graphs for IoT systems, showing all of the system resources that can be compromised and enumerating potential attack traces. In building Iota, we design novel techniques to scan IoT systems for individual vulnerabilities and further create generic exploit models for IoT vulnerabilities. We also identify and model physical dependencies between different devices as they are unique to IoT systems and are employed by adversaries to launch complicated attacks. In addition, we utilize NLP techniques to extract IoT app semantics based on app descriptions. To evaluate vulnerabilities' system-wide impact, we propose two metrics based on the attack graph, which provide guidance on fortifying IoT systems. Evaluation on 127 IoT CVEs (Common Vulnerabilities and Exposures) shows that Iota's exploit modeling module achieves over 80% accuracy in predicting vulnerabilities' preconditions and effects. We apply Iota to 37 synthetic smart home IoT systems based on real-world IoT apps and devices. Experimental results show that our framework is effective and highly efficient. Among 27 shortest attack traces revealed by the attack graphs, 62.8% are not anticipated by the system administrator. It only takes 1.2 seconds to generate and analyze the attack graph for an IoT system consisting of 50 devices.

cs.CR↗

arXiv2022

Blockchain Meets COVID-19: A Framework for Contact Information Sharing and Risk Notification System

COVID-19 is a severe global epidemic in human history. Even though there are particular medications and vaccines to curb the epidemic, tracing and isolating the infection source is the best option to slow the virus spread and reduce infection and death rates. There are three disadvantages to the existing contact tracing system: 1. User data is stored in a centralized database that could be stolen and tampered with, 2. User's confidential personal identity may be revealed to a third party or organization, 3. Existing contact tracing systems only focus on information sharing from one dimension, such as location-based tracing, which significantly limits the effectiveness of such systems. We propose a global COVID-19 information sharing and risk notification system that utilizes the Blockchain, Smart Contract, and Bluetooth. To protect user privacy, we design a novel Blockchain-based platform that can share consistent and non-tampered contact tracing information from multiple dimensions, such as location-based for indirect contact and Bluetooth-based for direct contact. Hierarchical smart contract architecture is also designed to achieve global agreements from users about how to process and utilize user data, thereby enhancing the data usage transparency. Furthermore, we propose a mechanism to protect user identity privacy from multiple aspects. More importantly, our system can notify the users about the exposure risk via smart contracts. We implement a prototype system to conduct extensive measurements to demonstrate the feasibility and effectiveness of our system.

cs.CR↗

arXiv2021

VRM-Phase I VKW system description of long-short video customizable keyword wakeup challenge

Keyword wakeup technology has always been a research hotspot in speech processing, but many related works were done on different datasets. We organized a Chinese long-short video keyword wakeup challenge (Video Keyword Wakeup Challenge, VKW) for testing the ability of each participating team to build a keyword wakeup system under the public dataset. All submitted systems not only need to support the setting of multiple different keywords, but also need to support the wakeup of any costumed keyword.This paper mainly describes the basic situation of the VKW challenge and the experimental results of some participating teams.

cs.SD↗

arXiv2021

Membership Inference Attacks Against Recommender Systems

Recently, recommender systems have achieved promising performances and become one of the most widely used web applications. However, recommender systems are often trained on highly sensitive user data, thus potential data leakage from recommender systems may lead to severe privacy problems. In this paper, we make the first attempt on quantifying the privacy leakage of recommender systems through the lens of membership inference. In contrast with traditional membership inference against machine learning classifiers, our attack faces two main differences. First, our attack is on the user-level but not on the data sample-level. Second, the adversary can only observe the ordered recommended items from a recommender system instead of prediction results in the form of posterior probabilities. To address the above challenges, we propose a novel method by representing users from relevant items. Moreover, a shadow recommender is established to derive the labeled training data for training the attack model. Extensive experimental results show that our attack framework achieves a strong performance. In addition, we design a defense mechanism to effectively mitigate the membership inference threat of recommender systems.

cs.CR↗

arXiv2021

Dual Synchronous Generator: Inertial Current Source based Grid-Forming Solution for VSC

In order to improve dynamic characteristics of the power system with high-proportion renewable energy sources (RESs), it is necessary for the voltage source converter (VSC), interfaces of RESs, to provide inertial and frequency regulation. In practical applications, VSCs are better to be controlled as a current source due to its weak overcurrent capacity. According to the characteristic, a dual synchronous theory is proposed to analyze the synchronization between current sources in this paper. Based on dual synchronous idea, a dual synchronous generator (DSG) control is applied in VSC to form inertial current source. In addition, a braking control is embedded in DSG control to improve the transient stability of VSC. Finally, experimental results verify the effectiveness of the theory and the control method.

eess.SY↗

arXiv2020

IoTGaze: IoT Security Enforcement via Wireless Context Analysis

Internet of Things (IoT) has become the most promising technology for service automation, monitoring, and interconnection, etc. However, the security and privacy issues caused by IoT arouse concerns. Recent research focuses on addressing security issues by looking inside platform and apps. In this work, we creatively change the angle to consider security problems from a wireless context perspective. We propose a novel framework called IoTGaze, which can discover potential anomalies and vulnerabilities in the IoT system via wireless traffic analysis. By sniffing the encrypted wireless traffic, IoTGaze can automatically identify the sequential interaction of events between apps and devices. We discover the temporal event dependencies and generate the Wireless Context for the IoT system. Meanwhile, we extract the IoT Context, which reflects user's expectation, from IoT apps' descriptions and user interfaces. If the wireless context does not match the expected IoT context, IoTGaze reports an anomaly. Furthermore, IoTGaze can discover the vulnerabilities caused by the inter-app interaction via hidden channels, such as temperature and illuminance. We provide a proof-of-concept implementation and evaluation of our framework on the Samsung SmartThings platform. The evaluation shows that IoTGaze can effectively discover anomalies and vulnerabilities, thereby greatly enhancing the security of IoT systems.

cs.CR↗

arXiv2020

Shielding Collaborative Learning: Mitigating Poisoning Attacks through Client-Side Detection

Collaborative learning allows multiple clients to train a joint model without sharing their data with each other. Each client performs training locally and then submits the model updates to a central server for aggregation. Since the server has no visibility into the process of generating the updates, collaborative learning is vulnerable to poisoning attacks where a malicious client can generate a poisoned update to introduce backdoor functionality to the joint model. The existing solutions for detecting poisoned updates, however, fail to defend against the recently proposed attacks, especially in the non-IID setting. In this paper, we present a novel defense scheme to detect anomalous updates in both IID and non-IID settings. Our key idea is to realize client-side cross-validation, where each update is evaluated over other clients' local data. The server will adjust the weights of the updates based on the evaluation results when performing aggregation. To adapt to the unbalanced distribution of data in the non-IID setting, a dynamic client allocation mechanism is designed to assign detection tasks to the most suitable clients. During the detection process, we also protect the client-level privacy to prevent malicious clients from stealing the training data of other clients, by integrating differential privacy with our design without degrading the detection performance. Our experimental evaluations on two real-world datasets show that our scheme is significantly robust to two representative poisoning attacks.

cs.CR↗