Deep Learning Methods in Speaker Recognition: A Review

Schiferl, D.; Szaszák, György; Beke, András

doi:10.3311/ppee.17024

Cited by 15 publications

(11 citation statements)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, we talk about open set (or out-of-set) speaker identification when the set of known speakers does not contain potential test subjects [7].…”

Section: Introductionmentioning

confidence: 99%

“…In speaker verification, we verify that the speaker is who he/ she says he/she is by comparing two (or more) speech samples/utterances and evaluating whether the speakers in the two samples are the same [7]. This is traditionally done, in general forensic voice comparison practice, by comparing the test sample or samples with the given speaker's sample or samples and a universal background model [8].…”

Section: Introductionmentioning

confidence: 99%

“…The closed set (or in‐set) scenario is when all speakers within a given set are known. On the other hand, we talk about open set (or out‐of‐set) speaker identification when the set of known speakers does not contain potential test subjects [7].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Effects of language mismatch in automatic forensic voice comparison using deep learning embeddings

Schiferl

Fejes

2023

Journal of Forensic Sciences

Self Cite

View full text Add to dashboard Cite

In forensic voice comparison, deep learning has become widely popular recently. It is mainly used to learn speaker representations, called embeddings or embedding vectors. Speaker embeddings are often trained using corpora mostly containing widely spoken languages. Thus, language dependency is an important factor in automatic forensic voice comparison, especially when the target language is linguistically very different from that the model is trained on. In the case of a low‐resource language, developing a corpus for forensic purposes containing enough speakers to train deep learning models is costly. This study aims to investigate whether a model pre‐trained on multilingual (mostly English) corpus can be used on a target low‐resource language (here, Hungarian), not represented by the model. Often multiple samples are not available from the offender (unknown speaker). Samples are therefore compared pairwise with and without speaker enrollment for suspect (known) speakers. Two corpora are used that were developed especially for forensic purposes and a third that is meant for traditional speaker verification. Speaker embedding vectors are extracted by the x‐vector and ECAPA‐TDNN techniques. Speaker verification was evaluated in the likelihood‐ratio framework. A comparison is made between the language combinations (modeling, LR calibration, and evaluation). The results were evaluated by Cllrmin and EER metrics. It was found that the model pre‐trained on a different language but on a corpus with a significant number of speakers can be used on samples with language mismatch. Sample duration and speaking style also seem to affect the performance.

show abstract

“…On the other hand, we talk about open set (or out-of-set) speaker identification when the set of known speakers does not contain potential test subjects [7].…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Effects of language mismatch in automatic forensic voice comparison using deep learning embeddings

Schiferl

Fejes

2023

Journal of Forensic Sciences

Self Cite

View full text Add to dashboard Cite

show abstract

“…Most of the research on speech processing nowadays is focused on the use of Artificial Neural Networks and Deep Learning. Deep Belief Networks (DBN) are widely used in speech recognition [18], [19]. X-vectors are considered today state of the art in speaker recognition [20].…”

Section: Related Researchmentioning

confidence: 99%

Speaker Diarization and Identification From Single Channel Classroom Audio Recordings Using Virtual Microphones

2022

View full text Add to dashboard Cite

Speaker identification in noisy audio recordings, specifically those from collaborative learning environments, can be extremely challenging. There is a need to identify individual students talking in small groups from other students talking at the same time. To solve the problem, we assume the use of a single microphone per student group without any access to previous large datasets for training. This dissertation proposes a method of speaker identification using crosscorrelation patterns associated to an array of virtual microphones, centered around the physical microphone. The virtual microphones are simulated by using approximate speaker geometry observed from a video recording. The patterns are constructed based on estimates

show abstract

“…However, non-visual cues such as voice represent also important information for person recognition, especially when the person is not in the visual field of view or when vision is limited. Deep learning techniques have proved to be a robust tool also to address speaker recognition, achieving state-of-the-art results [9]. Nonetheless, several issues arise when applying AI methodologies to visual and non-visual information gathered in HRI frameworks.…”

Section: Introductionmentioning

confidence: 99%

Cognitive architecture aided by working-memory for self-supervised multi-modal humans recognition

Gonzalez-Billandon¹,

Belgiovine²,

Sciutti³

et al. 2021

Preprint

View full text Add to dashboard Cite

The ability to recognize human partners is an important social skill to build personalized and long-term human-robot interactions, especially in scenarios like education, care-giving, and rehabilitation. Faces and voices constitute two important sources of information to enable artificial systems to reliably recognize individuals. Deep learning networks have achieved state-of-the-art results and demonstrated to be suitable tools to address such a task. However, when those networks are applied to different and unprecedented scenarios not included in the training set, they can suffer a drop in performance. For example, with robotic platforms in ever-changing and realistic environments, where always new sensory evidence is acquired, the performance of those models degrades. One solution is to make robots learn from their first-hand sensory data with selfsupervision. This allows coping with the inherent variability of the data gathered in realistic and interactive contexts. To this aim, we propose a cognitive architecture integrating low-level perceptual processes with a spatial working memory mechanism. The architecture autonomously organizes the robot's sensory experience into a structured dataset suitable for human recognition. Our results demonstrate the effectiveness of our architecture and show that it is a promising solution in the quest of making robots more autonomous in their learning process.

show abstract

Deep Learning Methods in Speaker Recognition: A Review

Cited by 15 publications

References 65 publications

Effects of language mismatch in automatic forensic voice comparison using deep learning embeddings

Effects of language mismatch in automatic forensic voice comparison using deep learning embeddings

Speaker Diarization and Identification From Single Channel Classroom Audio Recordings Using Virtual Microphones

Cognitive architecture aided by working-memory for self-supervised multi-modal humans recognition

Contact Info

Product

Resources

About