Speaker recognition

About: Speaker recognition is a research topic. Over the lifetime, 14990 publications have been published within this topic receiving 310061 citations.

Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognition

Abstract: Recently, speaker embeddings extracted from a speaker discriminative deep neural network (DNN) yield better performance than the conventional methods such as i-vector. In most cases, the DNN speaker classifier is trained using cross entropy loss with softmax. However, this kind of loss function does not explicitly encourage inter-class separability and intra-class compactness. As a result, the embeddings are not optimal for speaker recognition tasks. In this paper, to address this issue, three different margin based losses which not only separate classes but also demand a fixed margin between classes are introduced to deep speaker embedding learning. It could be demonstrated that the margin is the key to obtain more discriminative speaker embeddings. Experiments are conducted on two public text independent tasks: VoxCeleb1 and Speaker in The Wild (SITW). The proposed approach can achieve the state-of-the-art performance, with 25% ~ 30% equal error rate (EER) reduction on both tasks when compared to strong baselines using cross entropy loss with softmax, obtaining 2.238% EER on VoxCeleb1 test set and 2.761% EER on SITW core-core test set, respectively.

Hybrid Speech Recognition

Abstract: A hybrid speech recognition system uses a client-side speech recognition engine and a server-side speech recognition engine to produce speech recognition results for the same speech. An arbitration engine produces speech recognition output based on one or both of the client-side and server-side speech recognition results.

Recognition of conversational telephone speech using the JANUS speech engine

Abstract: Recognition of conversational speech is one of the most challenging speech recognition tasks to-date. While recognition error rates of 10% or lower can now be reached on speech dictation tasks over vocabularies in excess of 60,000 words, recognition of conversational speech has persistently resisted most attempts at improvements by way of the proven techniques to date. Difficulties arise from shorter words, telephone channel degradation, and highly disfluent and coarticulated speech. In this paper, we describe the application, adaptation, and performance evaluation of our JANUS speech recognition engine to the Switchboard conversational speech recognition task. Through a number of algorithmic improvements, we have been able to reduce error rates from more than 50% word error to 38%, measured on the offical 1996 NIST evaluation test set. Improvements include vocal tract length normalization, polyphonic modeling, label boosting, speaker adaptation with and without confidence measures, and speaking mode dependent pronunciation modeling.

On the vulnerability of automatic speaker recognition to spoofing attacks with artificial signals

Abstract: Automatic speaker verification (ASV) systems are increasingly being used for biometric authentication even if their vulnerability to imposture or spoofing is now widely acknowledged. Recent work has proposed different spoofing approaches which can be used to test vulnerabilities. This paper introduces a new approach based on artificial, tone-like signals which provoke higher ASV scores than genuine client tests. Experimental results show degradations in the equal error rate from 8.5% to 77.3% and from 4.8% to 64.3% for standard Gaussian mixture model and factor analysis based ASV systems respectively. These findings demonstrate the importance of efforts to develop dedicated countermeasures, some of them trivial, to protect ASV systems from spoofing.

Babble Noise: Modeling, Analysis, and Applications

Abstract: Speech babble is one of the most challenging noise interference for all speech systems. Here, a systematic approach to model its underlying structure is proposed to further the existing knowledge of speech processing in noisy environments. This paper establishes a working foundation for the analysis and modeling of babble speech. We first address the underlying model for multiple speaker babble speech - considering the number of conversations versus the number of speakers contributing to babble. Next, based on this model, we develop an algorithm to detect the range of the number of speakers within an unknown babble speech sequence. Evaluation is performed using 110 h of data from the Switchboard corpus. The number of simultaneous conversations ranges from one to nine, or one to 18 subjects speaking. A speaker conversation stream detection rate in excess of 80% is achieved with a speaker window size of plusmn1 speakers. Finally, the problem of in-set/out-of-set speaker recognition is considered in the context of interfering babble speech noise. Results are shown for test durations from 2-8 s, with babble speaker groups ranging from two to nine subjects. It is shown that by choosing the correct number of speakers in the background babble an overall average performance gain of 6.44% equal error rate can be obtained. This study represents effectively the first effort in developing an overall model for speech babble, and with this, contributions are made for speech system robustness in noise.