Speaker Diarization for Low-Resource Languages Through Wav2vec Fine-Tuning

Abdulhady Abas Abdullah; Kanar R. Tariq; Sara Azad Ahmed; Sarkhel H. Taher Karim; Tarik A. Rashid

REVIEW 5 major objections 5 minor 61 references

Speaker Diarization for Low-Resource Languages Through Wav2vec Fine-Tuning

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Fine-tuning Wav2Vec 2.0 on a small Kurdish corpus cuts diarization error by 7.2 points.

desk verdict First Wav2Vec 2.0 fine-tuning for Kurdish diarization with a purpose-built corpus, but two contradictory result sets make the headline 7.2% DER improvement unsupported as written. read the letter →

arxiv 2504.18582 v1 pith:ESC4ULYH submitted 2025-04-23 cs.SD cs.CLeess.AS

Abdulhady Abas Abdullah , Sarkhel H. Taher Karim , Sara Azad Ahmed , Kanar R. Tariq , Tarik A. Rashid This is my paper · ORCID

classification cs.SDcs.CLeess.AS

keywords SpeakerDiarizationKurdishSpeechProcessingWav2Vec2.0Self-SupervisedLearningTransferLow-ResourceLanguagesErrorRateClusterPurity

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Speaker diarization—splitting an audio stream by who is speaking—works well for high-resource languages but has been barely studied for Kurdish, which lacks labeled corpora and spans several dialects. This paper tries to close that gap by taking a pre-trained Wav2Vec 2.0 model and fine-tuning it on a small, purpose-built Kurdish audio corpus. On the paper's reported five-run evaluation, fine-tuning lowers the Diarization Error Rate from 22.8% to 15.6% (a 7.2-point drop) and raises cluster purity from 76.4% to 89.1% (a 12.7-point gain). If those numbers hold, the same transfer-learning recipe could make speaker separation practical for other under-resourced languages.

What carries the argument

The load-bearing object is Wav2Vec 2.0, a self-supervised speech encoder that learns representations from raw audio using a convolutional network plus a transformer context network. The paper fine-tunes it with a dual loss: cross-entropy for speaker classification and connectionist temporal classification (CTC) for aligning speaker transitions. Data preprocessing—noise reduction, normalization, segmentation, and augmentation with synthetic noise, pitch shifts, and speed changes—feeds the model cleaner Kurdish speech.

What would settle it

Run the pre-trained and fine-tuned Wav2Vec models through one fixed diarization pipeline—same segmentation, same clustering, same scoring tool—on a held-out Kurdish test set. If the pre-trained baseline scores near the paper's other reported baseline value of 15.2% DER rather than 22.8%, the claimed 7.2-point improvement evaporates; a score near 22.8% would support the claim.

Watch

Extended reading notes

Core claim

The paper's claim is that a self-supervised multilingual speech model, Wav2Vec 2.0, can be adapted to Kurdish speaker diarization with only a small curated dataset. The argument is that the pre-trained model already encodes general acoustic and phonetic structure, and fine-tuning on Kurdish audio with two losses—cross-entropy for assigning segments to speakers and CTC loss for aligning speaker boundaries in time—shifts those representations toward Kurdish. The reported evidence shows DER falling from 22.8% to 15.6%, cluster purity rising from 76.4% to 89.1%, and SNR rising from 12.5 dB to 18.7 dB, with standard deviations of ±0.5%, ±0.7%, and ±0.3 dB across five runs. The paper takes these gains as showing that transfer learning plus data augmentation can overcome the absence of large labeled Kurdish diarization datasets.

Load-bearing premise

The central claim depends on the pre-trained baseline being scored through the same diarization pipeline and test set as the fine-tuned model; if the baseline pipeline was weaker, the 7.2-point drop is not caused by fine-tuning.

Editorial extensions

If this is right

If the reported gains replicate, Kurdish media transcription, meeting analysis, and call-center speaker separation can be built without waiting for large annotated Kurdish corpora.
The combination of Wav2Vec 2.0 pretraining, a small curated dataset, and dual-loss fine-tuning becomes a template for other low-resource languages whose phonetics differ from the pretraining data.
Full fine-tuning outperforms few-shot, one-shot, and zero-shot variants, but all tuned variants beat the untuned baseline, so even a handful of labeled Kurdish examples appears to help.
The simultaneous drop in DER and rise in cluster purity implies the fine-tuned model produces both fewer wrong speaker labels and more coherent speaker clusters, not just a trade-off between the two.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The paper does not isolate how much of the gain comes from fine-tuning versus preprocessing; an ablation that keeps the audio pipeline fixed while varying only the model weights would separate those contributions.
The same recipe could be tested on other under-resourced languages with dialect variation and code-switching, such as comparing Sorani versus Kurmanji audio, to see whether transfer learning generalizes beyond Kurdish.
A useful extension would measure how DER and cluster purity scale with the size of the labeled Kurdish corpus; the few-shot/one-shot/zero-shot table suggests the curve may be steep, which would tell practitioners how much annotation is worth paying for.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The genuinely new thing here is the target language: no one has built a Kurdish speaker diarization benchmark, and the paper addresses a real gap. The authors assembled a purpose-built corpus (269 files, up to four speakers), document their augmentation and hyperparameter choices, and honestly list dataset limitations such as dialect coverage and overlapping speech. That is useful scaffolding for the community.

The problem is the experimental reporting. The paper contains two mutually exclusive evaluations of the same comparison. The prose around Figure 8 says baseline DER 15.2%, fine-tuned 8.0%; the table labeled Table 1 in Section 4.3 says 22.8% and 15.6%. Both have the same 7.2-point absolute gap, which is suspicious. Cluster purity and SNR also contradict each other: 76.4%/89.1% and 12.5/18.7 dB versus 78.5%/85.3% and 10.4/14.7 dB. So the paper does not pin down a single empirical effect. The abstract's 7.2% reduction could be read from either version, but the supporting numbers disagree.

The baseline pipeline is also undefined. We never learn what segmentation, embedding, clustering, or scoring was used for the pre-trained Wav2Vec baseline. If the baseline ran through a weaker pipeline, the improvement is meaningless. The claim of p<0.05 over five runs has no test statistic, only standard deviations. The dataset and code are not released, so there is no way to check.

The citation pattern is sloppy: [19] is a medical imaging paper, [26] is about dollar store vaccine distribution. There are duplicated section numbers and a stray line about a "logo" that looks like an editing artifact. These are not fatal by themselves, but they reinforce the impression that the manuscript was not carefully checked.

To be fair, the core idea is plausible: fine-tuning a self-supervised multilingual model on Kurdish should help. The authors deserve credit for identifying the gap and building a first corpus. But the central claim is load-bearing, and the paper's own data does not uniquely support it. This needs a major revision: reconcile the numbers, specify the baseline pipeline, and release the dataset or at least a detailed experimental protocol.

Who is this for? Researchers working on Kurdish speech technology and low-resource diarization. They would want to know about the corpus and the approach. But as a citable result, it is not trustworthy in current form. A serious editor should send it to peer review rather than desk reject, because the reported gap is large and the language is genuinely underserved; a reviewer can demand the fixes. I would not accept it as is, and I would not cite it until the numbers are reconciled.

Referee Report

5 major / 5 minor

Summary. The paper proposes fine-tuning Wav2Vec 2.0 on a custom Kurdish audio corpus for speaker diarization, reporting improvements in Diarization Error Rate (DER), cluster purity, and Signal-to-Noise Ratio (SNR) relative to a pre-trained Wav2Vec baseline. It describes the dataset, preprocessing and augmentation steps, a dual cross-entropy/CTC fine-tuning objective, hyperparameter choices, and results for full, few-shot, one-shot, and zero-shot settings. The central claim is that fine-tuning reduces DER by 7.2 percentage points and improves cluster purity by about 13 percentage points, with implications for low-resource speech technology.

Significance. If the headline results were reproducible, the paper would address a genuine gap: speaker diarization for Kurdish is understudied, and a dedicated Kurdish corpus plus a transfer-learning recipe would be a useful community resource. The paper also makes a plausible case that data augmentation and fine-tuning can help in low-resource settings. However, the current manuscript does not support the central claim: the evaluation is internally contradictory, the baseline diarization pipeline is unspecified, and no code or data are released. These issues prevent external verification and make the reported effect size uninterpretable as written.

major comments (5)

[§4.2, §4.3, Table 1] The paper reports two irreconcilable sets of results for the same comparison. Table 1 gives baseline DER 22.8% and fine-tuned DER 15.6%, cluster purity 76.4% and 89.1%, and SNR 12.5 dB and 18.7 dB. Section 4.2 and Figure 8 report baseline DER 15.2% and fine-tuned DER 8.0%, cluster purity 78.5% and 85.3%, and SNR 10.4 dB and 14.7 dB. Both sets share the same 7.2-percentage-point DER gap, so the abstract's '7.2%' does not uniquely refer to either experiment. The authors must identify which numbers are the actual evaluation and provide a single consistent set of results; as written, the central empirical claim is unsupported.
[§3.3, §4.2] The baseline to which the fine-tuned model is compared is never defined as a diarization system. A Wav2Vec encoder alone does not produce speaker diarization; one must specify voice activity detection, speaker segmentation, embedding extraction, clustering, handling of the number of speakers, and the scoring tool. None of these steps is described for either the baseline or the fine-tuned model. Without this information, the reported DER drop cannot be attributed to fine-tuning rather than to a difference in the evaluation pipeline.
[§4.1, Table 1] The claimed statistical significance is not substantiated. Section 4.1 states that 'the exact p-value of less than 0.05 was obtained for all analysed parameters' and that the DER decrease was 'confirmed statistically different in five independent runs, SD = ±0.5%', but no test statistic, degrees of freedom, confidence intervals, or baseline variance are reported. Table 1 provides only a single standard-deviation column without clarifying whether it applies to both conditions or how it was computed; this is insufficient support for an inferential claim.
[§3.3.2, Table 1] The metric definitions are too incomplete to verify the reported improvements. DER is defined with a simple sum over total speech time, but standard diarization scoring requires a collar, forgiveness, and overlap handling; cluster purity is described only as 'the proportion of correctly grouped segments to the total number of segments', which presupposes the clustering that is never specified. SNR is an audio-quality metric, not a diarization metric, so the SNR gain in Table 1 reflects preprocessing rather than fine-tuning and should not be presented as part of the model's diarization improvement.
[§3.1.1, §4.1] The evaluation rests entirely on a custom, unreleased Kurdish corpus whose composition is not described in sufficient detail. Section 3.1.1 gives only folder counts and a qualitative list of limitations; there is no information on total duration, exact number of speakers per file beyond folder labels, dialect distribution, speaker demographics, or annotation protocol, and the dataset is not made available. Because every result is obtained on this private test set, the findings cannot be checked or compared with any external benchmark.

minor comments (5)

[§1] The objectives list refers to 'Improve Word2Vec for Kurdish' and the text later alternates between Word2Vec and Wav2Vec; the paper is about Wav2Vec 2.0, so the terminology should be corrected throughout.
[§4.2] The relative DER decrease from 15.2% to 8.0% is 47.4%, not 50.3% as stated; please recalculate and report the correct value.
[Table 3] Table 3 contains unexplained columns (LP, IDA, RI) and an RI value of 0.0 for the baseline; these should be defined or removed.
[§2.2.3] There are duplicated subsection headings labeled 2.2.3, and reference [26] does not appear to support the claim about low-resource diarization; the references should be verified.
[§3.1.1] Several passages appear unfinished or out of place, such as the sentence beginning 'A logo is a provisional description'; the manuscript needs careful editing.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the fine-tuning improvement is an empirical comparison, with inconsistencies that are reproducibility issues rather than definitional or self-citation reductions.

full rationale

The paper's central claim (fine-tuning Wav2Vec 2.0 on a Kurdish corpus improves diarization error rate and cluster purity relative to a pre-trained baseline) is an empirical result obtained by training on one dataset and scoring on a held-out test subset. There is no step in Sections 3.3-4.3 where the reported improvement is imposed by definition: DER, cluster purity, and SNR are computed from predictions and ground truth, while the fine-tuning procedure (CE+CTC losses, learning rate 1e-5, batch size 16, 20 epochs) is described independently. The pre-trained baseline is a different checkpoint, not a re-labeled version of the fine-tuned output, so the comparison is not a fitted-input-call-prediction. The paper's self-citations ([7] and [12], both by Abdullah et al.) are used only to support background statements about Kurdish resource scarcity and prior Kurdish ASR work; the numerical improvement is not justified by those citations. The inconsistent baseline DER (22.8% in Table 1 vs. 15.2% in Section 4.2, with correspondingly different final values) is a serious reproducibility and protocol weakness, but it does not constitute circularity: both numbers are presented as empirical measurements, not as quantities forced by construction or by the authors' own prior results. Thus the derivation chain is self-contained in the sense relevant to circularity analysis.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The ledger lists standard experimental hyperparameters chosen by hand, domain assumptions about the dataset and metrics, and no invented entities. The paper's claim rests on the Kurdish corpus being correctly annotated and representative, and on the baseline comparison using the same pipeline; both are assumptions the paper does not fully support.

free parameters (7)

learning_rate = 1e-5
Chosen via pilot experiments; controls convergence of fine-tuning.
batch_size = 16
Chosen for hardware memory and gradient stability; affects training dynamics.
epochs = 20 with early stopping
Set to avoid overfitting on the small dataset.
weight_decay = 0.01
AdamW regularization strength chosen to stabilize fine-tuning.
augmentation_noise_intensity = 5% wide-band noise
Hand-set level for synthetic noise augmentation.
pitch_shift_range = pm 5 semitones
Hand-set range for pitch alteration augmentation.
speed_range = 0.9x to 1.1x
Hand-set range for speed modification augmentation.

assumptions (5)

domain assumption Wav2Vec 2.0 representations pretrained on multilingual data transfer to Kurdish acoustic and phonetic features
Section 3.4 invokes transfer learning; this is a background assumption about representation transfer.
domain assumption The custom Kurdish diarization corpus has accurate speaker labels and timestamps
Section 3.1.1 admits potential annotation issues; training and evaluation rely on label correctness.
domain assumption DER and cluster purity are computed according to the stated formulas
Section 3.3.2 defines metrics; no external scoring tool is cited or released.
domain assumption The 70/20/10 split is representative of Kurdish multi-speaker audio
Section 4.1 states the split; the dataset itself is small and dialect-limited.
domain assumption Statistical significance of the 7.2 point DER drop follows from five runs
Section 4.1 asserts p<0.05 without reporting the test or distribution; the claim depends on this unverified statistical assumption.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Speaker Diarization for Low-Resource Languages Through Wav2vec Fine-Tuning." pith.science (2026). https://pith.science/paper/ESC4ULYH

@misc{pith2026250418582,
  author       = {Pith},
  title        = {Pith review of: Speaker Diarization for Low-Resource Languages Through Wav2vec Fine-Tuning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ESC4ULYH}},
  note         = {Machine review of arXiv:2504.18582}
}

read the original abstract

Speaker diarization is a fundamental task in speech processing that involves dividing an audio stream by speaker. Although state-of-the-art models have advanced performance in high-resource languages, low-resource languages such as Kurdish pose unique challenges due to limited annotated data, multiple dialects and frequent code-switching. In this study, we address these issues by training the Wav2Vec 2.0 self-supervised learning model on a dedicated Kurdish corpus. By leveraging transfer learning, we adapted multilingual representations learned from other languages to capture the phonetic and acoustic characteristics of Kurdish speech. Relative to a baseline method, our approach reduced the diarization error rate by seven point two percent and improved cluster purity by thirteen percent. These findings demonstrate that enhancements to existing models can significantly improve diarization performance for under-resourced languages. Our work has practical implications for developing transcription services for Kurdish-language media and for speaker segmentation in multilingual call centers, teleconferencing and video-conferencing systems. The results establish a foundation for building effective diarization systems in other understudied languages, contributing to greater equity in speech technology.

Figures

Figures reproduced from arXiv: 2504.18582 by the authors.

**Figure 3.** Wav2Vec Model An instance of this may be seen in the XLSR model created by Conneau et al., which shown significant improvements in ASR for languages that have low resources. This enhancement was accomplished by first doing pre-training on 53 different languages, followed by fine-tuning the model particularly for the desired language. This model modification enables the transmission of extensive acoustic and linguist… view at source ↗

**Figure 4.** Fine tuning in Speaker Diarization The challenge of assessing speaker diarization in low-resource languages such as Kurdish is made more difficult by the limited availability of standardized benchmark datasets. Researchers often encounter the requirement to generate their own assessment sets, which makes it difficult to directly compare different research. Recent developments, as shown by [40], showcase innovative m… view at source ↗

**Figure 8.** Performance comparison between baseline and fine-tuned Models [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

61 extracted references · 53 canonical work pages

[19]

KuBERT: Central Kurdish BERT Model and Its Application for Sentiment Analysis,

H. Veisi, K. muhealddin Awlla, and A. A. Abdullah, "KuBERT: Central Kurdish BERT Model and Its Application for Sentiment Analysis," 2024

work page 2024
[26]

A survey on recent approaches for natural language processing in low -resource scenarios,

M. A. Hedderich, L. Lange, H. Adel, J. Strötgen, and D. Klakow, "A survey on recent approaches for natural language processing in low -resource scenarios," arXiv preprint arXiv:2010.12309, 2020

arXiv 2010
[1]

This work has gained significant importance in the field of speech processing

Introduction Speaker diarization is the act of dividing an audio stream into segments that are similar in terms of the speaker's identification, as illustrated in Figure 1 . This work has gained significant importance in the field of speech processing. Technology addresses the inquiry of identifying the individuals who spoke at certain times in speech rec...

work page
[2]

Improve Word2Vec for Kurdish: Refine the word2vec models by using a dataset labeled with Kurdish speaker diarization, hence enhancing the model's proficiency in identifying and analyzing Kurdish phonetics

work page
[3]

Contribute to the enhancement of robust Kurdish language models, particularly in the domain of voice recognition, with the aim of achieving higher accuracy and reliability

work page
[4]

Finally, Conclusion and Future Work

Advancing Multilingual NLP: This aims to promote inclusivity in NLP technology by tackling the difficulties encountered in low-represented languages such as Kurdish. The subsequent sections of this work are structured in the following manner: The following section, Related Work, examines the current body of research on speaker diarization, word2vec, and K...

work page
[5]

data augmentation

Related Work 2.1 Exploring Speaker Diarization: Methods and Innovations Diarization tasks, such as the identification of who spoke when , have improved over a period of time. The kind of approach, however, that is common involves several processes which are discrete, the first being speech activity detection (SAD), which refers to the identification of a ...

work page
[6]

The approach starts by providing a comprehensive depiction of the dataset, including its organization and the preprocessing procedures executed to make it suitable for training

Methodology This chapter presents the technique to refine the Wav2Vec model for speaker diarization using Kurdish audio data as explained in figure 5. The approach starts by providing a comprehensive depiction of the dataset, including its organization and the preprocessing procedures executed to make it suitable for training. Subsequently, a comprehensiv...

work page

Show all 61 references

[7]

Ensuring the model's ability to differentiate between distinct voices was crucial, especially for recordings involving many speakers [41]

Noise Reduction: Noise reduction methods were used to diminish background noise and extraneous noises, hence improving the clarity of the audio signals. Ensuring the model's ability to differentiate between distinct voices was crucial, especially for recordings involving many ...
[8]

By normalizing the data, the model is able to prioritize the distinct attributes of each speaker's voice, without being affected by differences in volume [42]

Normalisation: The audio files underwent normalization to guarantee uniform volume levels across the collection. By normalizing the data, the model is able to prioritize the distinct attributes of each speaker's voice, without being affected by differences in volume [42]
[9]

Segmentation aids in the training of the model to identify shifts in speakers and enhances its capacity to process lengthy audio re cordings [1]

Segmentation: The audio files were divided into smaller, more easily handled sections, particularly for those that had numerous speakers. Segmentation aids in the training of the model to identify shifts in speakers and enhances its capacity to process lengthy audio re cordings [1]
[10]

These strategies enhance the model's resilience to various acoustic circumstances and speaker varianc es [43]

Data Augmentation: Further data augmentation methods, such as introducing synthetic noise, altering pitch, or modifying speed, were used to enhance the dataset's diversity. These strategies enhance the model's resilience to various acoustic circumstances and speaker varianc es...
[11]

Synthetic Noise Addition: Wide-band noise (at 5% intensity level) and background conversation were added to the audio files. This improves generalization since real-world scenarios frequently present cases whereby speech overlaps with noise , and the method optimizes the model...
[12]

Pitch Alteration: The parameters of the speed/semantic rate were set at a range of ±5 semitones for the audio segments. This change is similar to modeling the natural variation of pitch intensity , making it easier for the model to learn speaker embeddings that are female and ...
[13]

This change really considers practical situations where speakers may speak at different tempos in order to enhance the model for variation in time

Speed Modification: Speech rate was also controlled at the range from 0.9x to 1.1x the original playback rate. This change really considers practical situations where speakers may speak at different tempos in order to enhance the model for variation in time. All the above proc...
[14]

The initial learning rate was fixed at 1e -5 as set by previous experiments and adjusted with a constant cosine rate to obtain convergence

Results and Discussion Specifically, the training configuration was explicitly set to achieve good performance while at the same time being computationally efficient. The initial learning rate was fixed at 1e -5 as set by previous experiments and adjusted with a constant cosin...
[15]

Conclusion This study showed that Wav2Vec 2.0 model fine -tuning succeeded at speaker diarization tasks within Kurdish , which counts as a low -resource language. The proposed approach delivered a minimal Diarization Error Rate (DER) decrease of 7.2% while raising cluster puri...
[16]

A review of speaker diarization: Recent advances with deep learning,

T. J. Park, N. Kanda, D. Dimitriadis, K. J. Han, S. Watanabe, and S. Narayanan, "A review of speaker diarization: Recent advances with deep learning," Computer Speech & Language, vol. 72, p. 101317, 2022

2022
[17]

End-to-end speaker diarization for an unknown number of speakers with encoder-decoder based attractors,

S. Horiguchi, Y. Fujita, S. Watanabe, Y. Xue, and K. Nagamatsu, "End-to-end speaker diarization for an unknown number of speakers with encoder-decoder based attractors," arXiv preprint arXiv:2005.09921, 2020

2005 arXiv
[18]

Language and Speech Technology for Central Kurdish Varieties,

S. Ahmadi, D. Q. Jaff, M. M. I. Alam, and A. Anastasopoulos, "Language and Speech Technology for Central Kurdish Varieties," arXiv preprint arXiv:2403.01983, 2024

2024 arXiv
[20]

wav2vec 2.0: A framework for self -supervised learning of speech representations,

A. Baevski, Y. Zhou, A. Mohamed, and M. Auli, "wav2vec 2.0: A framework for self -supervised learning of speech representations," Advances in neural information processing systems, vol. 33, pp. 12449-12460, 2020

2020
[21]

wav2vec: Unsupervised pre -training for speech recognition,

S. Schneider, A. Baevski, R. Collobert, and M. Auli, "wav2vec: Unsupervised pre -training for speech recognition," arXiv preprint arXiv:1904.05862, 2019

1904 arXiv
[22]

Breaking Walls: Pioneering Automatic Speech Recognition for Central Kurdish: End-to-End Transformer Paradigm,

A. A. Abdullah, H. Veisi, and T. Rashid, "Breaking Walls: Pioneering Automatic Speech Recognition for Central Kurdish: End-to-End Transformer Paradigm," arXiv preprint arXiv:2406.02561, 2024

2024 arXiv
[23]

Mls: A large-scale multilingual dataset for speech research,

V . Pratap, Q. Xu, A. Sriram, G. Synnaeve, and R. Collobert, "Mls: A large-scale multilingual dataset for speech research," arXiv preprint arXiv:2012.03411, 2020

2012 arXiv
[24]

Deep Learning for Natural Language Processing in Low -Resource Languages,

S. Tatineni, "Deep Learning for Natural Language Processing in Low -Resource Languages," International Journal of Advanced Research in Engineering and Technology (IJARET), vol. 11, no. 5, pp. 1301-1311, 2020

2020
[25]

A survey on text classification: From traditional to deep learning,

Q. Li et al., "A survey on text classification: From traditional to deep learning," ACM Transactions on Intelligent Systems and Technology (TIST), vol. 13, no. 2, pp. 1 -41, 2022

2022
[27]

Central Kurdish Automatic Speech Recognition using Deep Learning,

A. A. Abdullah and H. Veisi, "Central Kurdish Automatic Speech Recognition using Deep Learning," Journal of University of Anbar for Pure Science, vol. 16, no. 2, 2022

2022
[28]

Enhancing speaker diarization with large language models: A contextual beam search approach,

T. J. Park, K. Dhawan, N. Koluguri, and J. Balam, "Enhancing speaker diarization with large language models: A contextual beam search approach," in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , 2024, pp. 10861 -10865: IEEE

2024
[29]

2019 YEAR IN REVIEW: MACHINE LEARNING IN HEALTHCARE,

P . Mathur, A. K. Khanna, J. B. Cywinski, K. Maheshwari, D. F. Naylor Jr, and F. A. Papay, "2019 YEAR IN REVIEW: MACHINE LEARNING IN HEALTHCARE," Team BrainX, BrainX Community

2019
[30]

Speaker diarization: A review of recent research,

X. Anguera, S. Bozonnet, N. Evans, C. Fredouille, G. Friedland, and O. Vinyals, "Speaker diarization: A review of recent research," IEEE Transactions on audio, speech, and language processing, vol. 20, no. 2, pp. 356-370, 2012

2012
[31]

Approaches and applications of audio diarization,

D. A. Reynolds and P . Torres-Carrasquillo, "Approaches and applications of audio diarization," in Proceedings.(ICASSP'05). IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005., 2005, vol. 5, pp. v/953 -v/956 Vol. 5: IEEE

2005
[32]

Speaker diarization with PLDA i-vector scoring and unsupervised calibration,

G. Sell and D. Garcia -Romero, "Speaker diarization with PLDA i-vector scoring and unsupervised calibration," in 2014 IEEE Spoken Language Technology Workshop (SLT), 2014, pp. 413 -417: IEEE

2014
[33]

Speaker diarization with LSTM,

Q. Wang, C. Downey, L. Wan, P . A. Mansfield, and I. L. Moreno, "Speaker diarization with LSTM," in 2018 IEEE International conference on acoustics, speech and signal processing (ICASSP) , 2018, pp. 5239 -5243: IEEE

2018
[34]

FocusNet: imbalanced large and small organ segmentation with an end -to-end deep neural network for head and neck CT images,

Y. Gao et al., "FocusNet: imbalanced large and small organ segmentation with an end -to-end deep neural network for head and neck CT images," in Medical Image Computing and Computer Assisted Intervention–MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13 –...

2019
[35]

End -to-end neural speaker diarization with self-attention,

Y. Fujita, N. Kanda, S. Horiguchi, Y. Xue, K. Nagamatsu, and S. Watanabe, "End -to-end neural speaker diarization with self-attention," in 2019 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), 2019, pp. 296 -303: IEEE

2019
[37]

The third DIHARD diarization challenge,

N. Ryant et al., "The third DIHARD diarization challenge," arXiv preprint arXiv:2012.01477, 2020

2012 arXiv
[38]

Automatic speech recognition for under -resourced languages: A survey,

L. Besacier, E. Barnard, A. Karpov, and T. Schultz, "Automatic speech recognition for under -resourced languages: A survey," Speech communication, vol. 56, pp. 85-100, 2014

2014
[39]

Advances in Deep Speaker Verification: a study on robustness, portability, and security,

X. Liu, "Advances in Deep Speaker Verification: a study on robustness, portability, and security," Itä - Suomen yliopisto, 2023

2023
[40]

Towards end -to-end speaker diarization with generalized neural speaker clustering,

C. Zhang, J. Shi, C. Weng, M. Yu, and D. Yu, "Towards end -to-end speaker diarization with generalized neural speaker clustering," in ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2022, pp. 8372 -8376: IEEE

2022
[41]

Equity Impacts of Dollar Store Vaccine Distribution,

J. A. Chevalier, J. L. Schwartz, Y. Su, and K. R. Williams, "Equity Impacts of Dollar Store Vaccine Distribution," arXiv preprint arXiv:2104.01295, 2021

2021 arXiv
[42]

Kurdish interdialect machine translation,

H. Hassani, "Kurdish interdialect machine translation," in Proceedings of the fourth workshop on NLP for similar languages, varieties and dialects (VarDial) , 2017, pp. 63 -72

2017
[43]

Jira: a Central Kurdish speech recognition system, designing and building speech corpus and pronunciation lexicon,

H. Veisi, H. Hosseini, M. MohammadAmini, W . Fathy, and A. Mahmudi, "Jira: a Central Kurdish speech recognition system, designing and building speech corpus and pronunciation lexicon," Language Resources and Evaluation, vol. 56, no. 3, pp. 917 -941, 2022

2022
[44]

Kurdish dialect recognition using 1D CNN,

K. J. Ghafoor, K. M. H. Rawf, A. O. Abdulrahman, and S. H. Taher, "Kurdish dialect recognition using 1D CNN," ARO-The Scientific Journal of Koya University, vol. 9, no. 2, pp. 10 -14, 2021

2021
[45]

Effectiveness of self -supervised pre-training for asr,

A. Baevski and A. Mohamed, "Effectiveness of self -supervised pre-training for asr," in ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , 2020, pp. 7694 -7698: IEEE

2020
[46]

Exploring wav2vec 2.0 on speaker verification and language identification,

Z. Fan, M. Li, S. Zhou, and B. Xu, "Exploring wav2vec 2.0 on speaker verification and language identification," arXiv preprint arXiv:2012.06185, 2020

2012 arXiv
[47]

EEND-SS: Joint end-to-end neural speaker diarization and speech separation for flexible number of speakers,

S. Maiti et al., "EEND-SS: Joint end-to-end neural speaker diarization and speech separation for flexible number of speakers," in 2022 IEEE Spoken Language Technology Workshop (SLT) , 2023, pp. 480 -487: IEEE

2022
[48]

MSFNet: Multi-Scale Fusion Network for Brain -Controlled Speaker Extraction,

C. Fan et al., "MSFNet: Multi-Scale Fusion Network for Brain -Controlled Speaker Extraction," in ACM Multimedia 2024

2024
[49]

Unsupervised cross -lingual representation learning for speech recognition,

A. Conneau, A. Baevski, R. Collobert, A. Mohamed, and M. Auli, "Unsupervised cross -lingual representation learning for speech recognition," arXiv preprint arXiv:2006.13979, 2020

2006 arXiv
[50]

A survey on transfer learning,

S. J. Pan and Q. Yang, "A survey on transfer learning," IEEE Transactions on knowledge and data engineering, vol. 22, no. 10, pp. 1345-1359, 2009

2009
[51]

A survey on transfer learning in natural language processing,

Z. Alyafeai, M. S. AlShaibani, and I. Ahmad, "A survey on transfer learning in natural language processing," arXiv preprint arXiv:2007.04239, 2020

2007 arXiv
[52]

The NIST speaker recognition evaluation program,

A. Martin, M. Przybocki, and J. P . Campbell, "The NIST speaker recognition evaluation program," in Biometric Systems: Technology, Design and Performance Evaluation, J. Wayman, A. Jain, D. Maltoni, and D. Maio, Eds. London: Springer London, 2005, pp. 241 -262

2005
[53]

NSGA-II-DL: Metaheuristic optimal feature selection with Deep Learning Framework for HER2 classification in Breast Cancer,

J. Majidpour et al., "NSGA-II-DL: Metaheuristic optimal feature selection with Deep Learning Framework for HER2 classification in Breast Cancer," IEEE Access, 2024

2024
[54]

Diarization is Hard: Some Experiences and Lessons Learned for the JHU Team in the Inaugural DIHARD Challenge,

G. Sell et al., "Diarization is Hard: Some Experiences and Lessons Learned for the JHU Team in the Inaugural DIHARD Challenge," in Interspeech, 2018, pp. 2808 -2812

2018
[55]

Towards Unsupervised Speaker Diarization System for Multilingual Telephone Calls Using Pre -trained Whisper Model and Mixture of Sparse Autoencoders,

P . Lam, L. Pham, T. Nguyen, T. Pham, L. K. Nguyen, and A. Schindler, "Towards Unsupervised Speaker Diarization System for Multilingual Telephone Calls Using Pre -trained Whisper Model and Mixture of Sparse Autoencoders," arXiv preprint arXiv:2407.01963, 2024

2024 arXiv
[56]

Audacity (R): Free audio editor and recorder [Computer application]. Version 3.0. 0 retrieved March 17th, 2021,

A. Team, "Audacity (R): Free audio editor and recorder [Computer application]. Version 3.0. 0 retrieved March 17th, 2021," ed, 2021

2021
[57]

Praat: doing phonetics by computer [Computer program],

P . Boersma, "Praat: doing phonetics by computer [Computer program]," http://www. praat. org/, 2011

2011
[58]

Audio augmentation for speech recognition,

T. Ko, V . Peddinti, D. Povey, and S. Khudanpur, "Audio augmentation for speech recognition," in Interspeech, 2015, vol. 2015, p. 3586

2015
[59]

Improving language understanding by generative pre -training,

A. Radford, "Improving language understanding by generative pre -training," 2018

2018
[60]

Connectionist temporal classification: labelling unsegmented sequence data with recurrent neural networks,

A. Graves, S. Fernández, F. Gomez, and J. Schmidhuber, "Connectionist temporal classification: labelling unsegmented sequence data with recurrent neural networks," in Proceedings of the 23rd international conference on Machine learning, 2006, pp. 369 -376

2006
[61]

Topic segmentation with an aspect hidden Markov model,

D. M. Blei and P . J. Moreno, "Topic segmentation with an aspect hidden Markov model," in Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval, 2001, pp. 343 -348

2001
[62]

Dropout: a simple way to prevent neural networks from overfitting,

N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, "Dropout: a simple way to prevent neural networks from overfitting," The journal of machine learning research, vol. 15, no. 1, pp. 1929-1958, 2014

1929

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [19]

KuBERT: Central Kurdish BERT Model and Its Application for Sentiment Analysis,

H. Veisi, K. muhealddin Awlla, and A. A. Abdullah, "KuBERT: Central Kurdish BERT Model and Its Application for Sentiment Analysis," 2024

work page 2024

[2] [26]

A survey on recent approaches for natural language processing in low -resource scenarios,

M. A. Hedderich, L. Lange, H. Adel, J. Strötgen, and D. Klakow, "A survey on recent approaches for natural language processing in low -resource scenarios," arXiv preprint arXiv:2010.12309, 2020

arXiv 2010

[3] [1]

This work has gained significant importance in the field of speech processing

Introduction Speaker diarization is the act of dividing an audio stream into segments that are similar in terms of the speaker's identification, as illustrated in Figure 1 . This work has gained significant importance in the field of speech processing. Technology addresses the inquiry of identifying the individuals who spoke at certain times in speech rec...

work page

[4] [2]

Improve Word2Vec for Kurdish: Refine the word2vec models by using a dataset labeled with Kurdish speaker diarization, hence enhancing the model's proficiency in identifying and analyzing Kurdish phonetics

work page

[5] [3]

Contribute to the enhancement of robust Kurdish language models, particularly in the domain of voice recognition, with the aim of achieving higher accuracy and reliability

work page

[6] [4]

Finally, Conclusion and Future Work

Advancing Multilingual NLP: This aims to promote inclusivity in NLP technology by tackling the difficulties encountered in low-represented languages such as Kurdish. The subsequent sections of this work are structured in the following manner: The following section, Related Work, examines the current body of research on speaker diarization, word2vec, and K...

work page

[7] [5]

data augmentation

Related Work 2.1 Exploring Speaker Diarization: Methods and Innovations Diarization tasks, such as the identification of who spoke when , have improved over a period of time. The kind of approach, however, that is common involves several processes which are discrete, the first being speech activity detection (SAD), which refers to the identification of a ...

work page

[8] [6]

The approach starts by providing a comprehensive depiction of the dataset, including its organization and the preprocessing procedures executed to make it suitable for training

Methodology This chapter presents the technique to refine the Wav2Vec model for speaker diarization using Kurdish audio data as explained in figure 5. The approach starts by providing a comprehensive depiction of the dataset, including its organization and the preprocessing procedures executed to make it suitable for training. Subsequently, a comprehensiv...

work page