Article
Biochemical Research Methods
Joshua Peeples, Weihuang Xu, Romain Gloaguen, Diane Rowland, Alina Zare, Zachary Brym
Summary: This article proposes a new method called STARSEED, which incorporates spatial information through the Earth Mover's Distance (EMD) to compare root system distributions. The approach captures the response of sesame root systems for different genotypes and soil moisture levels. STARSEED can be generalized to other plants and provides valuable insights into root system architecture development and response to varying growth conditions. The code and data for the experiments are publicly available at: https://github.com/GatorSense/STARSEED.
Article
Computer Science, Information Systems
Zhihua Zhang, Xudong Xu, M. James C. Crabbe
Summary: This study improves the wavelet-thresholding techniques (DWT-H and DWT-S) by introducing the earth mover's distance (EMD) as the similarity measure for small-scale patches and incorporating joint bilateral filtering at higher noise levels.
Article
Computer Science, Artificial Intelligence
Tsz Nam Chan, Man Lung Yiu, Leong U. Hou
Summary: EMD is a robust measure of similarity between two histograms, widely used in various fields. Despite the computationally intensive nature, efficient upper and lower bound functions have been developed without error guarantee. This study focuses on computing an approximate EMD value with bounded error, proposing solutions that demonstrate efficiency and effectiveness through experimental results.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2021)
Article
Computer Science, Information Systems
Yanpeng Qu, Zheng Xu, Changjing Shang, Xiaolong Ge, Ansheng Deng, Qiang Shen
Summary: This paper proposes a robust attribute reduction algorithm based on Earth Mover's Distance (EMD), which optimizes the reduction results by denoising instances, leading to superior performance in terms of reduction size and classification results compared to other state-of-the-art techniques.
INFORMATION SCIENCES
(2021)
Article
Chemistry, Multidisciplinary
Longsheng Wei, Lei Zhao, Jian Peng
Summary: In this paper, a reduced reference quality assessment algorithm for image retargeting using earth mover's distance is proposed. Local and global feature information is extracted from reference and retargeted images, and a quality score is calculated based on the similarity measure using earth mover's distance. Experimental results show improvement in image quality scores for retargeted images.
APPLIED SCIENCES-BASEL
(2021)
Article
Quantum Science & Technology
Bobak Toussi Kiani, Giacomo De Palma, Milad Marvian, Zi-Wen Liu, Seth Lloyd
Summary: This paper introduces a solution to the problem of commonly used distance metrics in machine learning in quantum settings. It proposes a quantum EM distance as a quantum analog to the classical EM distance, which possesses unique properties that make quantum learning more stable and efficient. The paper also presents a quantum Wasserstein generative adversarial network (qWGAN) that takes advantage of the quantum EM distance for learning on quantum data.
QUANTUM SCIENCE AND TECHNOLOGY
(2022)
Article
Physiology
Vasanth Ravikumar, Sanket Thakare, Xiangzhen Kong, Henri Roukoz, Elena G. Tolkacheva
Summary: The study aimed to develop a similarity score using various iEGM analysis techniques to accurately identify the spatial location of active sites of arrhythmia in patients with AF.
FRONTIERS IN PHYSIOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Pattaramanee Arsomngern, Cheng Long, Supasorn Suwajanakorn, Sarana Nutanong
Summary: Deep metric learning is a supervised learning paradigm that constructs a meaningful vector space for representing complex objects. Its successful application to pointsets can eliminate expensive retrieval operations on objects and greatly facilitate various machine learning and data mining tasks involving pointsets.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Computer Science, Interdisciplinary Applications
Peixian Liang, Yizhe Zhang, Yifan Ding, Jianxu Chen, Chinedu S. Madukoma, Tim Weninger, Joshua D. Shrout, Danny Z. Chen
Summary: The study introduces a novel framework, hierarchical earth mover's distance (H-EMD), for instance segmentation in biomedical images, which effectively explores the probability maps generated by DL models to achieve better instance segmentation results.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Multidisciplinary Sciences
Sangwon Hyun, Aditya Mishra, Christopher L. Follett, Bror Jonsson, Gemma Kulk, Gael Forget, Marie-Fanny Racault, Thomas Jackson, Stephanie Dutkiewicz, Christian L. Muller, Jacob Bien
Summary: The Wasserstein distance is a powerful metric for analyzing marine ecosystems and climate predictions. It quantifies spatial displacement differences and successfully isolates temporal and depth variability, as well as quantifies shifts in biogeochemical province boundaries. Using the Wasserstein distance, we can better understand the temporal dynamics of the ocean and test models.
PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCIENCES
(2022)
Article
Physics, Multidisciplinary
Xiangzhen Kong, Vasanth Ravikumar, Siva K. K. Mulpuru, Henri Roukoz, Elena G. G. Tolkacheva
Summary: This article presents a data-driven preprocessing framework for analyzing intracardiac electrogram data from patients with atrial fibrillation. The study optimizes the parameters of a bandpass filter using a data-driven approach and demonstrates the effects on subsequent frequency analysis. The results show that a bandpass threshold of 15 Hz has the best performance.
Article
Mathematics
William Q. Erickson
Summary: The paper considers two natural statistics on pairs of histograms and derives formulas for the probability of EMD=|D| and the expected value of |D| using the combinatorics of Young diagrams and plane partitions.
DISCRETE MATHEMATICS
(2024)
Article
Remote Sensing
Lin Chen, Franz Rottensteiner, Christian Heipke
Summary: In feature based image matching, distinctive features in images are detected and represented by feature descriptors. Matching is then carried out by assessing the similarity of the descriptors of potentially conjugate points. The evolution from hand-crafted feature descriptors, e.g. SIFT, to machine learning and deep learning based descriptors is discussed in detail, along with the advantages and challenges of different approaches.
GEO-SPATIAL INFORMATION SCIENCE
(2021)
Article
Computer Science, Information Systems
Rui Feng, Haitao Dong, Xuri Li, Zhaochuang Gu, Runyang Tian, Houde Li
Summary: An improved support vector machine with earth mover's distance (EMD-SVM) is proposed, which considers the data structure information and automatically learns the distribution between classes. Experimental validation demonstrates its superior and robust performance.
Article
Geochemistry & Geophysics
Jiaxing Sun, Xiaobo Shen, Quansen Sun
Summary: This article proposes a novel meta-learning method for hyperspectral image (HSI) few-shot classification. It utilizes a few labeled samples for classification. The Earth mover's distance (EMD) is introduced as a metric, and the EMD metric learning module is designed to calculate the similarity of embedding features. The proposed method outperforms existing HSI methods according to extensive experimental results.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Information Systems
Hongwei Luo, Yijie Shen, Feng Lin, Guoai Xu
Summary: This paper introduces an attack strategy against speaker verification systems based on deep neural networks, with a success rate of up to 82% and high imperceptibility, capable of deceiving the system while remaining hidden from human hearing or machine discrimination.
SECURITY AND COMMUNICATION NETWORKS
(2021)
Article
Computer Science, Information Systems
Aditya Singh Rathore, Chenhan Xu, Weijin Zhu, Afee Daiyan, Kun Wang, Feng Lin, Kui Ren, Wenyao Xu
Summary: This paper introduces a new fingerprint sensing technology called SonicPrint, which utilizes the intrinsic fingerprint ridge information in sonic wave for user identification. It is a practical technology that requires no hardware modifications and leverages the built-in microphones in smart devices. The experimental results show a high accuracy rate.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2022)
Article
Computer Science, Information Systems
Huining Li, Chenhan Xu, Aditya Singh Rathore, Zhengxiong Li, Hanbin Zhang, Chen Song, Kun Wang, Lu Su, Feng Lin, Kui Ren, Wenyao Xu
Summary: With the increasing use of voice-controlled devices, voice metrics have become popular for user identification. However, voice biometrics are susceptible to replay attacks and ambient noise. In this paper, the authors present VocalPrint, a mmWave interrogation system that directly captures and analyzes vocal vibrations for user authentication. By exploiting the disturbance in radio frequency signals caused by vocal vibrations around the near-throat area, VocalPrint is able to isolate ambient noise and preserve fine-grained vocal biometric properties. Experimental results demonstrate the resilience of VocalPrint against complex noise interference and spoof attacks, achieving over 96 percent authentication accuracy even under unfavorable conditions.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Computer Science, Hardware & Architecture
Jianwei Liu, Kaiyan Cui, Xiang Zou, Jinsong Han, Feng Lin, Kui Ren
Summary: In this paper, the authors propose a secure and user-friendly multi-factor user authentication system called BioDraw. It utilizes four categories of biometrics and a pattern-based password for user identification and authentication. The system achieves high authentication accuracy and is effective in defending against various attacks, as demonstrated through extensive experiments.
IEEE-ACM TRANSACTIONS ON NETWORKING
(2023)
Proceedings Paper
Computer Science, Theory & Methods
Meng Chen, Li Lu, Jiadi Yu, Yingying Chen, Zhongjie Ba, Feng Lin, Kui Ren
Summary: Faced with the threat of identity leakage during voice data publishing, users are engaged in a privacy-utility dilemma while enjoying convenient voice services. Existing studies employ direct modification or text-based re-synthesis to de-identify users' voices, but resulting in inconsistent audibility for human participants and not adaptive to informed attacks. In this poster, we propose a non-intrusive and adaptive speaker de-identification scheme to balance the privacy and utility of voice services.
PROCEEDINGS OF THE 2022 THE 28TH ANNUAL INTERNATIONAL CONFERENCE ON MOBILE COMPUTING AND NETWORKING, ACM MOBICOM 2022
(2022)
Proceedings Paper
Computer Science, Theory & Methods
Chao Wang, Feng Lin, Tiantian Liu, Kaidi Zheng, Zhibo Wang, Zhengxiong Li, Ming-Chun Huang, Wenyao Xu, Kui Ren
Summary: This study presents a remote attack on smartphone earpieces using mmWave sensors to eavesdrop on emitted speech. Through optimizing the fitting function and denoising scheme, the attack range can be extended to 6-8m, posing a threat to 23 different models of smartphones.
PROCEEDINGS OF THE 2022 THE 28TH ANNUAL INTERNATIONAL CONFERENCE ON MOBILE COMPUTING AND NETWORKING, ACM MOBICOM 2022
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Chao Wang, Feng Lin, Tiantian Liu, Ziwei Liu, Yijie Shen, Zhongjie Ba, Li Lu, Wenyao Xu, Kui Ren
Summary: This paper presents mmPhone, a novel acoustic eavesdropping system that recovers loudspeaker speech protected by soundproof environments. By utilizing mmWaves to sense the piezoelectric film and decoding the speech, mmPhone is capable of recovering high-quality and intelligible speech from a certain distance.
IEEE CONFERENCE ON COMPUTER COMMUNICATIONS (IEEE INFOCOM 2022)
(2022)
Article
Computer Science, Information Systems
Zhibo Wang, Defang Liu, Yunan Sun, Xiaoyi Pang, Peng Sun, Feng Lin, John C. S. Lui, Kui Ren
Summary: With the development of communication technologies and IoT infrastructures, home automation systems have become popular for providing convenient smart-home services. However, there are security risks in the deployment and application of these systems, which require research on attack and defense. This paper presents a comprehensive survey on the security of home automation systems, including system architecture, attack classification, vulnerability analysis, security requirements, existing defense methods, and open issues for future research.
IEEE COMMUNICATIONS SURVEYS AND TUTORIALS
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Yike Chen, Ming Gao, Yimin Li, Lingfeng Zhang, Li Lu, Feng Lin, Jinsong Han, Kui Ren
Summary: This paper investigates the ability of ultrasonic microphone jammers (UMJs) to resist covert eavesdropping and proposes a comprehensive framework for evaluating the resilience of UMJs. Extensive assessment results reveal that most existing UMJs are vulnerable to sophisticated adverse approaches, and suggestions for future designs of UMJs are presented.
IEEE CONFERENCE ON COMPUTER COMMUNICATIONS (IEEE INFOCOM 2022)
(2022)
Article
Computer Science, Information Systems
Chao Wang, Feng Lin, Zhongjie Ba, Fan Zhang, Wenyao Xu, Kui Ren
Summary: Most existing eavesdropping attacks that rely on sound waves for speech retrieval are ineffective in scenarios protected by soundproof materials. This paper presents a novel method using portable and commercial off-the-shelf mmWave devices to compromise human speech in isolated rooms, along with a word detection system called Wavesdropper that utilizes a mmWave probe to sense throat vibrations and recover speech contents. Experimental results show the effectiveness and accuracy of Wavesdropper in speech recognition.
PROCEEDINGS OF THE ACM ON INTERACTIVE MOBILE WEARABLE AND UBIQUITOUS TECHNOLOGIES-IMWUT
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Li Lu, Zhongjie Ba, Feng Lin, Jinsong Han, Kui Ren
Summary: This paper introduces ActListener, a method that eavesdrops on user activities using WiFi infrastructure without their knowledge. The proposed attack does not require physical access to the user's device or prior knowledge of activity recognition models and device locations. Experimental results show that ActListener achieves good performance in recovering original signals and activity recognition.
2022 IEEE 42ND INTERNATIONAL CONFERENCE ON DISTRIBUTED COMPUTING SYSTEMS (ICDCS 2022)
(2022)
Proceedings Paper
Biophysics
Kaidi Zheng, Chao Wang, Zhanglei Shu, Feng Lin
Summary: This paper proposes a novel speech acquisition method that directly senses sound waves through the air instead of relying on sound-induced vibrations. The study finds that mmWave signals can penetrate soundproofing materials, allowing for speech recovery through obstacles. To combat the loss of mmWave signals, a speech enhancement method based on deep neural networks is developed.
2022 IEEE MTT-S INTERNATIONAL MICROWAVE BIOMEDICAL CONFERENCE (IMBIOC)
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Jianwei Liu, Xiang Zou, Feng Lin, Jinsong Han, Xian Xu, Kui Ren
Summary: This paper introduces Hand-Key, a novel user authentication system that utilizes RFID technology to collect both inner and outer body features for user identification, addressing the challenge of balancing user-friendliness and security in biometric-based authentication. By leveraging the inherent randomness of RFID systems, Hand-Key is immune to replay attacks and achieves an authentication success rate of 99%+.
2021 IEEE 41ST INTERNATIONAL CONFERENCE ON DISTRIBUTED COMPUTING SYSTEMS (ICDCS 2021)
(2021)
Proceedings Paper
Telecommunications
Liu Liu, Hanlin Yu, Zhongjie Ba, Li Lu, Feng Lin, Kui Ren
Summary: Facial recognition technology plays a crucial role in mobile authentication, but faces various impersonation attacks. To address this issue, this paper proposes a new anti-spoofing facial recognition system, PassFace, which uses raw facial videos as the second factor for authentication to enhance security and resist attacks. Experiment results show that PassFace can achieve satisfactory performance in authentication and attack resistance.
IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC 2021)
(2021)
Proceedings Paper
Computer Science, Hardware & Architecture
Weiye Xu, Jianwei Liu, Shimin Zhang, Yuanqing Zheng, Feng Lin, Jinsong Han, Fu Xiao, Kui Ren
Summary: The research introduces a novel privacy-preserving anti-spoofing FA system called RFace, which extracts both 3D geometry and inner biomaterial features of faces using a COTS RFID tag array. Experimental results show that RFace has a high authentication success rate and is not deceived by any spoofing attacks.
IEEE CONFERENCE ON COMPUTER COMMUNICATIONS (IEEE INFOCOM 2021)
(2021)