Speaker Recognition With Random Digit Strings Using Uncertainty Normalized HMM-Based i-Vectors
| Autor: | Nooshin Maghsoodi, Hossein Zeinali, Hossein Sameti, Themos Stafylakis |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
FOS: Computer and information sciences
Sound (cs.SD) Computer Science - Computation and Language Acoustics and Ultrasonics Channel (digital image) Computer science Speech recognition Feature extraction Frame (networking) Word error rate Speech processing Speaker recognition Computer Science - Sound Bottleneck 030507 speech-language pathology & audiology 03 medical and health sciences Computational Mathematics Audio and Speech Processing (eess.AS) FOS: Electrical engineering electronic engineering information engineering Computer Science (miscellaneous) Electrical and Electronic Engineering 0305 other medical science Hidden Markov model Computation and Language (cs.CL) Electrical Engineering and Systems Science - Audio and Speech Processing |
| Zdroj: | IEEE/ACM Transactions on Audio, Speech, and Language Processing. 27:1815-1825 |
| ISSN: | 2329-9304 2329-9290 |
| Popis: | In this paper, we combine Hidden Markov Models (HMMs) with i-vector extractors to address the problem of text-dependent speaker recognition with random digit strings. We employ digit-specific HMMs to segment the utterances into digits, to perform frame alignment to HMM states and to extract Baum-Welch statistics. By making use of the natural partition of input features into digits, we train digit-specific i-vector extractors on top of each HMM and we extract well-localized i-vectors, each modelling merely the phonetic content corresponding to a single digit. We then examine ways to perform channel and uncertainty compensation, and we propose a novel method for using the uncertainty in the i-vector estimates. The experiments on RSR2015 part III show that the proposed method attains 1.52\% and 1.77\% Equal Error Rate (EER) for male and female respectively, outperforming state-of-the-art methods such as x-vectors, trained on vast amounts of data. Furthermore, these results are attained by a single system trained entirely on RSR2015, and by a simple score-normalized cosine distance. Moreover, we show that the omission of channel compensation yields only a minor degradation in performance, meaning that the system attains state-of-the-art results even without recordings from multiple handsets per speaker for training or enrolment. Similar conclusions are drawn from our experiments on the RedDots corpus, where the same method is evaluated on phrases. Finally, we report results with bottleneck features and show that further improvement is attained when fusing them with spectral features. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|