Speech enhancement has substantial interest in the utilization of speaker identification,
video-conference, speech transmission through communication channels, speech-based …

This paper develops an explainable artificial intelligence (AI) approach to measuring service
quality in voice-based service encounters. Drawing from the psychology and computer …

Abstract Language Identification (LI) is one of the widely emerging field in the areas of
speech processing to accurately identify the language from the data base based on some …

With the recent research developments, deep learning models are powerful alternatives for
speech enhancement and recognition in many real-world applications. Although state-of-the …

Speech recognition systems, trained and updated based on large-scale audio data, are
vulnerable to backdoor attacks that inject dedicated triggers in system training. The used …

Acoustic eavesdrop** on private or confidential spaces poses a significant privacy threat.
While soundproof rooms can mitigate these risks to some extent, they are unable to counter …

Earphones have become a popular voice input and interaction device. However, airborne
speech is susceptible to ambient noise, making it necessary to improve the quality and …

This paper discusses the problem of single channel speech enhancement in stationary
environments, and proposes Wiener filtering with the recursive noise estimation algorithm …

Speech enhancement algorithms aim to improve the quality and intelligibility of noise
corrupted speech, through spectral or temporal modifications. Most of the existing speech …

Speech enables easy human-to-human communication as well as human-to-machine
interaction. However, the quality of speech degrades due to background noise in the …