China
Africa
Explore chapters and articles related to this topic
Cybersecurity in Post COVID-19 Digital Era
Published in Kenneth Okereafor, Cybersecurity in the COVID-19 Pandemic, 2021
For example, an AI-enabled Cybersecurity using automatic speaker recognition technology could be optimized to match a voice to a person [12] for investigating high-profile cybercrime incidents involving telecommunications fraud, drug trafficking, kidnapping, and cyber espionage. This could also be applied for counterterrorism and predictive analytics of pre-ransomware attack stages by using neural networks to design software algorithms that listen to the isolated keywords in voice patterns and are able to predict suspected malicious plans with high-precision assurances.
Long term average speech spectra of Turkish
Published in Logopedics Phoniatrics Vocology, 2018
Differences have long been observed between acoustic properties of different languages, with many studies reporting on these variations. Acoustic properties of spoken languages and the effect on these features by variables such as gender and age are well established (1–5). Fundamental frequency, frequency and amplitude perturbations, and noise-to-harmonic ratio are parameters that are frequently used in evaluations, particularly when related to voice disorders. Voice analysis methods are also needed to create and measure hearing aid fitting algorithms and optimize automatic speaker recognition applications (1,6). Long-term average spectra of speech (LTASS) offer more general and inclusive information to meet these aims (7,8). To obtain LTASS, the frequency and amplitude distribution are analyzed following a long speech sample recording. This generates an acoustic representation of the language in daily conservations, allowing the researcher to obtain additional information on spectral energy distribution of a speech signal in a longer speech sample.
Novel navigation assistive device for deaf drivers
Published in Assistive Technology, 2022
Mwaffaq Otoom, Mohammad A. Alzubaidi, Rama Aloufee
Voice recognition is divided into two main types; speaker recognition and speech recognition. Speaker recognition focuses on identifying the person who is speaking (Campbell, 1997). Speech recognition focuses on distinguishing and identifying the words that are spoken (Rabiner, Juang, & Rutledge, 1993). In this work, we focus on speech recognition, as the identity of the speaker is not important.