An overview of robustness related issues in speaker recognition

Abstract

Speaker recognition technologies have been improved rapidly in recent years. However, critical robustness issues need to be addressed when they are applied in practical situations. This paper provides an overview of technologies dealing with robustness related issues in automatic speaker recognition. We first categorize the robustness issues into three categories, including environment-related, speaker-related and application-oriented issues. For each category, we then describe the current hot topics, existing technologies, and potential research focuses in the future.I.

Deep speaker verification: Do we need end to end?

“…After decades of research, SV has gained significant performance improvement, and has been deployed in some practical applications [1], [2], [3], [4]. However, SV is still a very challenging task, mainly attributed to the large uncertainty caused by the complex convolution of various speech factors, especially phone content and channel [5].…”

Automatic speaker recognition from speech signal using bidirectional long‐short‐term memory recurrent neural network

“…10 The advancement in ASR systems for real-world applications is a challenging task. There are numerous robust issues based on the intended applications available, which make the ASR system vulnerable 11 ; noise and channel variations are two common issues in the ASR system. Most of the state-of-art techniques employ speech enhancement and noise removal techniques to reduce noise contamination.…”

Infected Phonemes: How a Cold Impairs Speech on a Phonetic Level

“…These symptoms directly affect the vocal tract causing a temporary speech disorder. Such speaker-related variability are one of the main difficulties in speech signal processing [7]. Yet, due to the lack of adequate speech databases, research on cold speech is still rare.…”

“…Therefore, the corpus provided for the Cold sub-challenge containing audio recordings of 630 German speakers with a total duration of approximately 45 hours offers an excellent "playground" to study how a cold influences a speaker's voice. The gained knowledge can help improving the robustness of speech related tasks such as speech recognition, speaker identification [7] or emotion recognition [8].…”