scholarly journals Identification of Fake Stereo Audio Using SVM and CNN

Information ◽  
2021 ◽  
Vol 12 (7) ◽  
pp. 263
Author(s):  
Tianyun Liu ◽  
Diqun Yan ◽  
Rangding Wang ◽  
Nan Yan ◽  
Gang Chen
Keyword(s):  
The Other ◽  
Perceptual Quality ◽  
Support Vector ◽  
Digital Audio ◽  
Audio Quality ◽  
Vector Machines ◽  
Mel Frequency Cepstral Coefficient ◽  
Audio System ◽  
Haas Effect ◽  
Effect Technique

The number of channels is one of the important criteria in regard to digital audio quality. Generally, stereo audio with two channels can provide better perceptual quality than mono audio. To seek illegal commercial benefit, one might convert a mono audio system to stereo with fake quality. Identifying stereo-faking audio is a lesser-investigated audio forensic issue. In this paper, a stereo faking corpus is first presented, which is created using the Haas effect technique. Two identification algorithms for fake stereo audio are proposed. One is based on Mel-frequency cepstral coefficient features and support vector machines. The other is based on a specially designed five-layer convolutional neural network. The experimental results on two datasets with five different cut-off frequencies show that the proposed algorithm can effectively detect stereo-faking audio and has good robustness.

scholarly journals MULTIPLE AUTHORS DETECTION: A QUANTITATIVE ANALYSIS OF DREAM OF THE RED CHAMBER

2014 ◽  
Vol 06 (04) ◽  
pp. 1450012 ◽  
Author(s):  
XIANFENG HU ◽  
YANG WANG ◽  
QIANG WU
Keyword(s):  
Machine Learning ◽  
Quantitative Analysis ◽  
Relative Frequency ◽  
Strong Support ◽  
The Other ◽  
Authorship Attribution ◽  
Support Vector ◽  
Feature Ranking ◽  
Vector Machines ◽  
Cao Xueqin

Inspired by the authorship controversy of Dream of the Red Chamber and the application of machine learning in the study of literary stylometry, we develop a rigorous new method for the mathematical analysis of authorship by testing for a so-called chrono-divide in writing styles. Our method incorporates some of the latest advances in the study of authorship attribution, particularly techniques from support vector machines. By introducing the notion of relative frequency as a feature ranking metric, our method proves to be highly effective and robust. Applying our method to the Cheng–Gao version of Dream of the Red Chamber has led to convincing if not irrefutable evidence that the first 80 chapters and the last 40 chapters of the book were written by two different authors. Furthermore, our analysis has unexpectedly provided strong support to the hypothesis that Chapter 67 was not the work of Cao Xueqin either. We have also tested our method to the other three Great Classical Novels in Chinese. As expected no chrono-divides have been found. This provides further evidence of the robustness of our method.

scholarly journals Automatic Classification of Adventitious Respiratory Sounds: A (Un)Solved Problem?

Sensors ◽  
2020 ◽  
Vol 21 (1) ◽  
pp. 57
Author(s):  
Bruno Machado Rocha ◽  
Diogo Pessoa ◽  
Alda Marques ◽  
Paulo Carvalho ◽  
Rui Pedro Paiva
Keyword(s):  
Automatic Classification ◽  
The Other ◽  
Support Vector ◽  
Respiratory Sounds ◽  
Linear Discriminant ◽  
Complex Evaluation ◽  
Vector Machines ◽  
Negative Class ◽  
Respiratory Conditions

(1) Background: Patients with respiratory conditions typically exhibit adventitious respiratory sounds (ARS), such as wheezes and crackles. ARS events have variable duration. In this work we studied the influence of event duration on automatic ARS classification, namely, how the creation of the Other class (negative class) affected the classifiers’ performance. (2) Methods: We conducted a set of experiments where we varied the durations of the other events on three tasks: crackle vs. wheeze vs. other (3 Class); crackle vs. other (2 Class Crackles); and wheeze vs. other (2 Class Wheezes). Four classifiers (linear discriminant analysis, support vector machines, boosted trees, and convolutional neural networks) were evaluated on those tasks using an open access respiratory sound database. (3) Results: While on the 3 Class task with fixed durations, the best classifier achieved an accuracy of 96.9%, the same classifier reached an accuracy of 81.8% on the more realistic 3 Class task with variable durations. (4) Conclusion: These results demonstrate the importance of experimental design on the assessment of the performance of automatic ARS classification algorithms. Furthermore, they also indicate, unlike what is stated in the literature, that the automatic classification of ARS is not a solved problem, as the algorithms’ performance decreases substantially under complex evaluation scenarios.

scholarly journals Pengenalan Karakter Plat Nomor Kendaraan Bermotor Menggunakan Zoning dan Fitur Freeman Chain Code

2018 ◽  
Vol 14 (1) ◽  
pp. 19-25
Author(s):  
Taufik Fuadi Abidin ◽  
Abbas Adam AzZuhri ◽  
Fitri Arnia
Keyword(s):  
Support Vector Machines ◽  
Character Recognition ◽  
Classification Problem ◽  
The Other ◽  
Support Vector ◽  
License Plate ◽  
Chain Code ◽  
Area Code ◽  
Vector Machines ◽  
Multi Class Classification

A license plate is one of the vehicle identities. It consists of alphabetic characters and numbers and represents provincial and area code where the vehicle is registered. This article discusses the character recognition of plate number using zoning and Freeman Chain Code (FCC). Zoning divides character image into several zones i.e. 4, 6, and 8, and then, the pattern of each character in the zone is extracted using FCC as the numerical features. The character is then classified using Support Vector Machines (SVM). It is a multi-class classification problem with 36 categories. The results show that FCC features with 8 zones give the best accuracy (87%) when compared to the other two zones.

Automatic Voice-Based Recognition For Automotive Headlights Beam Control

2021 ◽  
Vol 18 (1) ◽  
Author(s):  
W. Astuti ◽  
S. Tan ◽  
M.I. Solihin ◽  
R.S. Vincent ◽  
B. Michael
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Recognition System ◽  
Support Vector ◽  
Recognition Mechanism ◽  
Training Time ◽  
Performance Accuracy ◽  
Vector Machines ◽  
Mel Frequency Cepstral Coefficient ◽  
The Voice

Driving comfort plays an important role in modern automotive technologies. One of the ways of comforting the driver is the voice-based recognition to control car headlights. The driver uttered a ‘specific word’ that is taken as an input to the proposed voice-based recognition system. The proposed mechanism then determines if the signal was either ‘high beam’ or ‘low beam’ to control the car headlights. To activate the headlight’s beam, this voice recognised signal is sent to a processing board. Mel Frequency Cepstral Coefficient (MFCC) is used in the recognition mechanism to extract the uttered word before being fed into Artificial Neural Networks (ANN) and Support Vector Machines (SVM) as a classification engine. The proposed automatic voice-based recognition was evaluated via experimental work. The results show that the proposed automatic voice-based recognition for headlights activation control involving MFCC feature works effectively in which SVM gives slightly better performance accuracy when compared to ANN. In addition to a lesser training time, the resulting accuracy using SVM in the training and testing phase is 93.595% and 91.74% respectively. Meanwhile, ANN has an accuracy of 89.39% and 88.16% in the training and testing respectively.

Design of 5.1 Channel Audio System Based on Zigbee and PurePath Wireless Technology

2013 ◽  
Vol 385-386 ◽  
pp. 1296-1299
Author(s):  
De Sheng Wang ◽  
Zhi Chao Wang ◽  
Cheng Wu Lin
Keyword(s):  
Ieee 802.15.4 ◽  
Low Cost ◽  
Low Complexity ◽  
System Management ◽  
Protocol Stack ◽  
Digital Audio ◽  
Audio Quality ◽  
Zigbee Protocol ◽  
Low Efficiency ◽  
Audio System

This paper presents a design scheme of 5.1-channel wireless audio system which takes the advantages of Zigbee and PurePath Wireless technologies. It overcomes several disadvantages of traditional wired audio system such as inconvenience of speaker placement, complexity of wiring, low efficiency, etc. This system adopts a star topology structure with a host and six slave machines. System management is based on Zigbee network, and digital audio is transmitted via PurePath Wireless audio link. The CC8530 audio transceiver chips and IEEE 802.15.4 RF chips were adopted as the hardware platform and the Zigbee protocol stack was adopted as the core software. The results show that this method makes the wireless audio system low-cost, low-power and low-complexity. In addition, audio quality and system efficiency are both greatly enhanced.

scholarly journals Identifying super-feminine, super-masculine and sex-defining connections in the human braingraph

2021 ◽  
Author(s):  
László Keresztes ◽  
Evelin Szögi ◽  
Bálint Varga ◽  
Vince Grolmusz
Keyword(s):  
Female Subject ◽  
Male Subject ◽  
The Other ◽  
Support Vector ◽  
Present Contribution ◽  
Human Connectome Project ◽  
Vector Machines ◽  
The Subject ◽  
The Right ◽  
Brain Connections

AbstractFor more than a decade now, we can discover and study thousands of cerebral connections with the application of diffusion magnetic resonance imaging (dMRI) techniques and the accompanying algorithmic workflow. While numerous connectomical results were published enlightening the relation between the braingraph and certain biological, medical, and psychological properties, it is still a great challenge to identify a small number of brain connections closely related to those conditions. In the present contribution, by applying the 1200 Subjects Release of the Human Connectome Project (HCP) and Support Vector Machines, we identify just 102 connections out of the total number of 1950 connections in the 83-vertex graphs of 1064 subjects, which—by a simple linear test—precisely, without any error determine the sex of the subject. Next, we re-scaled the weights of the edges—corresponding to the discovered fibers—to be between 0 and 1, and, very surprisingly, we were able to identify two graph edges out of these 102, such that, if their weights are both 1, then the connectome always belongs to a female subject, independently of the other edges. Similarly, we have identified 3 edges from these 102, whose weights, if two of them are 1 and one is 0, imply that the graph belongs to a male subject—again, independently of the other edges. We call the former 2 edges superfeminine and the first two of the 3 edges supermasculine edges of the human connectome. Even more interestingly, the edge, connecting the right Pars Triangularis and the right Superior Parietal areas, is one of the 2 superfeminine edges, and it is also the third edge, accompanying the two supermasculine connections if its weight is 0; therefore, it is also a “switching” edge. Identifying such edge-sets of distinction is the unprecedented result of this work.

scholarly journals Radius Margin Bounds for Support Vector Machines with the RBF Kernel

2003 ◽  
Vol 15 (11) ◽  
pp. 2643-2681 ◽  
Author(s):  
Kai-Min Chung ◽  
Wei-Chun Kao ◽  
Chia-Liang Sun ◽  
Li-Lun Wang ◽  
Chih-Jen Lin
Keyword(s):  
The Other ◽  
Support Vector ◽  
Good Prediction ◽  
Support Vectors ◽  
Svm Model ◽  
Vector Machines ◽  
Comparable Performance ◽  
Important Approach ◽  
Rbf Kernel ◽  
Leave One Out

An important approach for efficient support vector machine (SVM) model selection is to use differentiable bounds of the leave-one-out (loo) error. Past efforts focused on finding tight bounds of loo (e.g., radius margin bounds, span bounds). However, their practical viability is still not very satisfactory. Duan, Keerthi, and Poo (2003) showed that radius margin bound gives good prediction for L2-SVM, one of the cases we look at. In this letter, through analyses about why this bound performs well for L2-SVM, we show that finding a bound whose minima are in a region with small loo values may be more important than its tightness. Based on this principle, we propose modified radius margin bounds for L1-SVM (the other case) where the original bound is applicable only to the hard-margin case. Our modification for L1-SVM achieves comparable performance to L2-SVM. To study whether L1-or L2-SVM should be used, we analyze other properties, such as their differentiability, number of support vectors, and number of free support vectors. In this aspect, L1-SVM possesses the advantage of having fewer support vectors. Their implementations are also different, so we discuss related issues in detail.

The Application of HIWO–SVM in Analog Circuit Fault Diagnosis

2015 ◽  
Vol 29 (08) ◽  
pp. 1550019 ◽  
Author(s):  
Hongzhi Hu ◽  
Shulin Tian ◽  
Qing Guo ◽  
Aijia Ouyang
Keyword(s):  
Fault Diagnosis ◽  
Analog Circuit ◽  
The Other ◽  
Support Vector ◽  
De Algorithm ◽  
Svm Algorithm ◽  
Vector Machines ◽  
Diagnosis Model ◽  
Test Points ◽  
Circuit Fault Diagnosis

The paper proposes a fault diagnosis model based on the HIWO–SVM algorithm given the fact that the basic support vector machines (SVM) cannot solve effectively the problem of fault diagnosis in analog circuit. First of all, the wavelet package technique is adopted for extracting the information of the faults from the test points in the analog circuit. The differential evolution (DE) algorithm is then integrated with the purpose of improving the performance of the basic IWO algorithm, i.e. a hybrid IWO (HIWO) algorithm. The HIWO algorithm is further used to optimize the parameters of SVM in order to avoid the randomness of the parameter selection, thereby improving the diagnosis precision and robustness. The experimental results on a filter circuit show that the method is more effective and reliable than the other methods for fault diagnosis.

Sign in / Sign up

Export Citation Format

Share Document