Pith. sign in

REVIEW 5 major objections 5 minor 136 references

Robust Assamese Speech Recognition through Controlled Fine-Tuning of Whisper Models

T0 review · 5 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read The paper claims that controlled fine-tuning of Whisper-Small on 3,520 Assamese clips—most of them unvalidated crowd audio—cuts word error from over 200 percent to 43.75 percent and curbs hallucination almost entirely.

desk verdict A useful empirical data point for Whisper-small on Assamese, but the central novelty—unvalidated data as regularization—is untested due to a missing validated-only control. read the letter →

arxiv 2607.17164 v1 pith:COUZ4WUO submitted 2026-07-19 cs.LG

Ganapati Das , Dwipen Laskar , Hasin Afzal Ahmed , Sanjib Kr Kalita , Kshirod Sarmah , Hem Chandra Das , Manjula Kalita This is my paper
classification cs.LG
keywords AutomaticSpeechRecognitionAssameseWhisperFine-tuningLow-resourcelanguagesHallucinationWordErrorRateCommonVoice
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

The paper argues that a precisely controlled fine-tuning of Whisper-Small on 3,520 clips of Common Voice Assamese (953 validated plus 2,567 community-unvalidated clips) lifts the model from a state of total failure—WER above 2, BLEU of 0, hallucinated Devanagari or repetitive characters—to a working recognizer with WER 43.75%, CER 13.18%, and BLEU 30.81. The authors attribute the turnaround not to data volume but to a specific preparation pipeline: NFKC normalization plus a custom Assamese/Bengali grapheme mapping, and the deliberate inclusion of noisy unvalidated clips as natural regularization. They also credit a stable optimization recipe (mixed precision, gradient accumulation, warmup with cosine decay, weight decay, early stopping) and beam search decoding. If this holds, it gives low-resource languages a cheap, repeatable path to adapting huge multilingual speech models without curated corpora or high-end GPUs.

What carries the argument

The central object is Whisper-Small, a 244-million-parameter encoder-decoder transformer pre-trained on 680,000 hours of multilingual audio. The adaptation mechanism is a controlled fine-tuning pipeline whose critical components are: (1) NFKC Unicode normalization plus a custom Assamese/Bengali grapheme mapping to collapse equivalent spellings and prevent vocabulary fragmentation; (2) data augmentation by merging the validated training clips with 2,567 crowd-sourced clips that lack community validation, treating the acoustic and transcript noise as natural regularization; (3) an optimization recipe using FP16 mixed precision with gradient accumulation on a 16GB GPU, a small batch, linear war

What would settle it

Search the 2,567 unvalidated training clips for exact or near-duplicate transcripts of the 394 test clips after normalization; any overlap directly inflates the WER gain. Also, re-run the fine-tuning with several random seeds on a random 90/10 split of the full data and check whether the 43.75% WER reproduces.

Watch

Extended reading notes

Core claim

The paper's central claim is that Whisper-Small, which scores a near-total failure on Assamese in zero-shot evaluation (WER 2.0127, CER 1.9091, BLEU 0, and a predicted hallucination rate of 0.5552), can be turned into a functional Assamese ASR by a controlled fine-tuning protocol. Trained on 3,520 clips (953 validated plus 2,567 unvalidated 'other' clips) from the Common Voice 24.0 Assamese corpus, the fine-tuned model reaches WER 0.4375, CER 0.1318, BLEU 30.81, and hallucination rate 0.0183. The authors attribute the gains to transcript canonicalization (NFKC normalization plus a custom mapping that unifies Assamese and Bengali graphemes), to using the noisy unvalidated clips as a regulariz

Load-bearing premise

The load-bearing assumption is that the 394-clip test set is clean and representative of Assamese speech, and that none of the 2,567 unvalidated training clips overlaps with it; if these fail, the reported improvements are inflated.

Editorial extensions

If this is right

  • A new benchmark for Whisper-Small on Assamese: WER 43.75%, CER 13.18%, BLEU 30.81 on the Common Voice 24.0 test set.
  • Unvalidated crowd-sourced clips, normally ignored, can be used as a regularizer and enlarge the usable training set nearly fourfold (from 953 to 3,520 clips).
  • The gap between CER and WER (30.5 percentage points) shows the model has learned Assamese phonetics but not its morphology and word boundaries; the paper concludes that adding an explicit language model is the next step.
  • The hallucination rate (prediction-normalized) falls from 0.5552 to 0.0183, a 96.7% relative reduction, implying output is now grounded in the audio.
  • Real-time factor improves from 0.2943 to 0.1990, meaning the fine-tuned model transcribes faster than real time even on a 16GB consumer GPU.

Reading between the lines

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

  • The same recipe—NFKC normalization, grapheme unification, and mixing unvalidated clips—could likely transfer to other low-resource languages in the Common Voice corpus, since the steps are mostly script-agnostic; this is an extension the authors do not test.
  • The reported figures come from a single fine-tuning run evaluated on a 394-clip test set, so run-to-run variance is unknown; a multi-seed repetition or a k-fold split of the combined validated and unvalidated data would clarify how stable the 78% relative WER gain is.
  • The paper's own error analysis suggests that a lightweight morphological segmenter or explicit language model at the decoding stage might close the CER-WER gap faster than scaling to a larger Whisper variant.
  • The proposed Hallucination Error Rate (HER) definition, counting unaligned predicted tokens, is simple and could be adopted as a standard reporting metric for low-resource ASR, where hallucination is often the dominant failure mode.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper fine-tunes Whisper-Small on the Mozilla Common Voice 24.0 Assamese corpus, using the 953 validated training clips plus 2,567 unvalidated clips from other.tsv. The fine-tuned model is compared to a zero-shot baseline on a 394-clip test set, with reported improvements in WER (2.0127→0.4375), CER (1.9091→0.1318), MER, WIL, BLEU, METEOR, F1, hallucination error rate, and RTF. The paper also provides a qualitative analysis of orthographic and morphological errors in the Assamese output. The claimed contribution is a 'controlled fine-tuning' pipeline and the strategic use of unvalidated crowd-sourced data as a regularization mechanism.

Significance. If the results are reproducible, the paper would provide a useful benchmark for Whisper-Small on Assamese and a practical recipe for low-resource ASR with limited compute. The qualitative error analysis is informative, and the huge reduction in hallucination from a near-zero-shot state is a promising sign. However, the paper's distinctive claim — that unvalidated data helps as a regularizer — is never tested by an ablation, and several reported numbers are internally inconsistent. As it stands, the evidence supports the weaker claim that fine-tuning on any moderately sized Assamese corpus, validated or not, improves over zero-shot. That weaker claim is credible but not novel enough on its own to justify the headline contribution.

major comments (5)
  1. [Section 4.3 and Section 7] The central contribution, 'Strategic use of unvalidated crowd-sourced speech data for data augmentation,' is unsupported. Section 4.3 states that other.tsv clips 'can serve as a natural regularisation mechanism,' but Section 7 compares only the model trained on 3,520 clips (953 validated + 2,567 other) against zero-shot. There is no control model trained on the 953 validated clips alone. Without this ablation, the reported WER/CER/HER improvements cannot be attributed to the unvalidated data; they could be due entirely to fine-tuning, and the other.tsv data might even be harmful. This is directly testable and should be reported before the paper's central claim is made.
  2. [Abstract vs. Section 7.1.2 and Table 7] The reported WER is inconsistent: the abstract states 43.17%, Section 7.1.2 says 'a word error rate (WER) of 43.75%,' Table 7 reports 0.4375, and Table 5 shows 0.4407 at step 3000 on the dev set. The relative improvement of 78.26% is computed from 0.4375. The authors must reconcile these numbers and state exactly which split and checkpoint were used for the final test-set evaluation.
  3. [Section 7.1 and Table 7] All evaluation results are single-run point estimates on a 394-clip test set, with no error bars, bootstrap confidence intervals, or multiple seeds. Given the very small test set and the extreme zero-shot baseline, the stability of the claimed WER=0.4375, BLEU=30.81, and HER=0.0183 is unknown. At minimum, the authors should report bootstrap confidence intervals or results over at least three fine-tuning runs.
  4. [Section 7.1.3 and Table 7 (RTF)] The RTF improvement (0.2943 → 0.1990) is presented as a 32.38% efficiency gain, but no explanation is given for why fine-tuning should change inference speed, and no measurement methodology is provided (e.g., number of runs, hardware warm-up, batch size, averaging across clips). On a shared Kaggle T4 GPU, RTF is likely to be dominated by environmental variance; as reported, this metric is not reliable evidence of efficiency improvement.
  5. [Section 5.10 (HER) and Table 7] The Hallucination Error Rate definitions in Eqs. (14) and (15) depend on a Levenshtein alignment between PRED and REF. For the zero-shot baseline, many predictions are in Devanagari or other scripts, so a character-level alignment may not meaningfully identify 'hallucinated tokens.' Also, HER_ref can exceed 1, and its relationship to the insertion component of WER is not discussed. The authors should validate HER against a standard measure, or at least clarify its behavior on cross-script outputs.
minor comments (5)
  1. [Throughout] Typos and inconsist dent text: 'Word Infomation Loss' in the abstract, 'Smaple 276' in Section 7.2.3, and 'InDespite' in Section 7.1.1. These should be corrected.
  2. [Table 6] The column headers 'HR_Pred' and 'HR_Ref' are inconsistent with the notation HER_pred and HER_ref used elsewhere; ensure consistent symbols.
  3. [Section 7.1.2] The text says validation loss reached its lowest value at 2500 steps, but the final model is evaluated at 3000 steps. Clarify the checkpoint selection rule (early stopping patience of 2 would normally stop at 2500 or 3000?) and report the actual checkpoint used for test evaluation.
  4. [Section 4.2.2] The custom Assamese/Bengali grapheme mapping is described qualitatively but no mapping table or vocabulary details are given. Since this mapping directly affects WER/CER computations, a concrete specification would improve reproducibility.
  5. [Section 4.1] The relationship among validated.tsv, other.tsv, and the official train/dev/test splits is clear in general, but the exact count '2,567 unique clips from other.tsv' should be justified; the total clip count and the validated split imply a different number for other.tsv unless some clips are excluded. A short data-preprocessing table would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported improvements are empirical comparisons on an external held-out Common Voice split, not reductions to the model's own inputs.

full rationale

The paper's derivation chain is an empirical benchmark comparison: Whisper-small is evaluated zero-shot on the Common Voice 24.0 Assamese test.tsv split, then fine-tuned on train.tsv plus other.tsv, and evaluated again on the same held-out test.tsv. The reported improvements (WER 2.0127 to 0.4375, CER 1.9091 to 0.1318, BLEU 0 to 30.81, HER_pred 0.5552 to 0.0183) are computed with standard metrics or with HER, a transparently defined insertion-style hallucination count. No parameter is fitted to test.tsv, and no evaluation metric is defined in terms of the training objective or a fitted quantity, so the improvement is not forced by construction. The claim that unvalidated other.tsv clips provide a natural regularization mechanism is not supported by a validated-only control condition; that is an evidentiary/ablation gap, not circularity. The author self-citations (e.g., Bharali & Kalita 2021) appear only in the related-work survey and are not load-bearing for the central result. No step can be quoted where the output reduces to the input by definition or where a fitted value is renamed as a prediction.

Assumptions & free parameters 7 free parameters · 6 assumptions · 1 invented entities

The central results rest on the Common Voice transcriptions, the untested inclusion of unvalidated clips, and a self-defined hallucination metric; hyperparameters are chosen by hand and not varied. No new physical entities are introduced; the HER metric is a new evaluative construct with no external validation.

free parameters (7)
  • peak learning rate = 1e-05
    Chosen by hand (Section 6.3.2); not tuned by search.
  • warmup steps = 50
    Chosen by hand (Section 6.3.2).
  • weight decay = 0.05
    Chosen by hand to regularize 244M-parameter model on small data (Section 6.3.2).
  • dropout rate = 0.05
    Chosen by hand (Section 6.3.2).
  • maximum training steps = 3000
    Fixed with early stopping patience 2 (Section 6.3.2, Table 2).
  • beam width = 5
    Decoding hyperparameter chosen by hand (Section 6.4).
  • number of unvalidated clips added = 2567
    Data-selection choice for augmentation (Section 4.1); not ablated.
assumptions (6)
  • domain assumption Common Voice 24.0 transcripts are accurate ground truth for training and evaluation.
    Invoked throughout Sections 4 and 7; if the crowd-sourced transcripts contain errors, metrics are miscalibrated.
  • ad hoc to paper Unvalidated other.tsv clips are acoustically valid Assamese and their inclusion acts as natural regularization.
    Stated in Section 4.3; not tested by ablation.
  • domain assumption NFKC normalization plus a custom Assamese-to-Bengali grapheme mapping preserves the linguistic identity of the transcript.
    Section 4.2.2; if the mapping changes phonetic content, the model is trained on corrupted labels.
  • domain assumption Levenshtein alignment is a valid operationalization of hallucination for Assamese.
    Section 5.10 defines HER using minimum edit distance; no external validation.
  • domain assumption The 394-clip test split is representative and independent of the training data.
    Section 4.1; no speaker-overlap analysis or confidence intervals are provided.
  • ad hoc to paper Whisper-Small is the appropriate model capacity: smaller variants cannot capture Assamese phonetics, larger variants overfit.
    Asserted in Section 3 without experimental evidence.
invented entities (1)
  • Hallucination Error Rate (HER_pred, HER_ref)
    purpose: Quantify hallucinated tokens in ASR output via Levenshtein alignment
    New metric defined in Section 5.10; not validated against human judgments or existing hallucination benchmarks.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Assamese Speech Recognition through Controlled Fine-Tuning of Whisper Models." pith.science (2026). https://pith.science/paper/COUZ4WUO

@misc{pith2026260717164,
  author       = {Pith},
  title        = {Pith review of: Robust Assamese Speech Recognition through Controlled Fine-Tuning of Whisper Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/COUZ4WUO}},
  note         = {Machine review of arXiv:2607.17164}
}
read the original abstract

Developing Automatic Speech Recognition (ASR) for morphologically rich, low-resource languages such as Assamese is challenging due to insufficient annotated speech data. The pretrained Whisper model performs poorly on Assamese speech recognition tasks. This paper presents a controlled, fine-tuned Whisper-based Assamese ASR system trained on the Mozilla Common Voice 24.0-Assamese corpus. A hardware-aware optimized training pipeline is implemented for resource-constrained environments, employing mixed-precision training and gradient accumulation on Tesla 4 Graphics Processing Units (T4 GPUs). The proposed fine-tuned model significantly outperformed the Zero-shot baseline, yielding Word Error Rate (WER), Character Error Rate (CER), Match Error Rate (MER), and Word Infomation Loss (WIL) of 43.17\%, 13.18\%, 43\%, and 64.81\%, respectively, achieving significant relative improvements of 78.26\%, 93.10\%, 57.0\%, and 35.19\% over the baseline. Semantic evaluation of the fine-tuned model also demonstrates notable improvement over a zero baseline, attaining Bilingual Evaluation Understudy (BLEU) and Metric for Evaluation of Translation with Explicit ORdering (METEOR) scores of 30.81 and 0.5262, respectively. Additionally, the predicted hallucination rate and Real-Time Factor (RTF) are substantially improved by 96.70\% and 32.38\%, compared to the zero-shot baseline.

Figures

Figures reproduced from arXiv: 2607.17164 by the authors.

Figure 1
Figure 1. Timeline of Automatic Speech Recognition History (1877-2025) (iii) Establishment of a performance benchmark for Whisper-Small in Assamese speech recognition. (iv) Qualitative analysis of orthographic and morphological errors, highlighting key challenges in Assamese ASR. The remainder of this paper is organized as follows. Section 2 provides a review of related works in ASR, with a focus on models and datasets used f… view at source ↗
Figure 2
Figure 2. Whisper model architecture. development (dev.tsv), and 394 test (test.tsv) samples. In addition to the official 953 validated training clips (train.tsv), the model was trained on 2,567 unique clips from the ‘other.tsv’ increasing the training set to 3520 clips. 4.2. Data Preprocessing Data preprocessing is a crucial step in fine-tuning any Acoustic model. Whisper requires both the audio signal and its corresponding … view at source ↗
Figure 3
Figure 3. Pre-processed Audio required by Whisper model for "আমাৰ ঘৰত তলা মৰা অৱস্াত ওছৰেৰ ঘৰত মানহ নথকা সময়ত ু চুিৰ ৈহিছল।".([amaôgH OôOt tOla mOôa ObOxtHat usOôOôe gH OôOt manuH nOtH Oka xOmOjOt suôi Hoisil]) Short-Time Fourier Transform (STFT) to a linear-frequency spectrogram. The Mel spectrogram shown in the third plot represents the converted audio signals in perceptually meaningful Mel scale bands that mimics human aud… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Zero-shot, Fine-tuning, and Testing Methodology. 𝑌ˆ = arg max 𝑌 Ë< ∏𝑛 𝑖=1 𝑃 (𝑦𝑖 Ý 𝑦<𝑖, 𝑋; 𝜃) (18) The discrepancy between the predicted sequence 𝑌ˆ and the ground-truth 𝑌𝑟𝑒𝑓 Ë 𝑡𝑎𝑟𝑔𝑒𝑡 is quantified using the WER, CER, MER, BLEU Score, METEOR, WIL, Precision, Recall, F…
Figure 5
Figure 5. Figure 5: Whisper-Small fine-tuning results for Assamese showing training and validation loss progression [PITH_FULL_IMAGE:figures/full_fig_p020_5.png]
Figure 6
Figure 6. Figure 6: Whisper-Small Fine-Tuning Results (Assamese) (a) Precision, Recall, F1-score Metrics (b) Hallucination Metrics [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]
Figure 7
Figure 7. Figure 7: Whisper-Small Fine-Tuning Results (Assamese) 7.1.4. Evaluation Metric Analysis Fine-tuned Whisper-small produces a significant improvement over the zero-shot evaluation with WER and CER of 43.75% and 13.18%, respectively. However, a substantial 30.51% gap remains betwe…
Figure 8
Figure 8. Figure 8: Performance Comparison of Zero-Shot (baseline) and Fine-tuned Whisper-Small model. character in a word is predicted incorrectly, the entire word is counted as wrong. As a result, minor spelling mistakes can significantly increase the WER value. In contrast, Character E…
Figure 9
Figure 9. Figure 9: Performance Comparison of Zero-Shot (baseline) and Fine-tuned Whisper-Small model. 8. These samples correspond to the same reference script used for the zero-shot evaluation of the model presented in [PITH_FULL_IMAGE:figures/full_fig_p023_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

136 extracted references · 25 canonical work pages

  1. [1]

    Journal of Information and Data Management , volume =

    Evaluation of automatic speech recognition approaches , author =. Journal of Information and Data Management , volume =. 2022 , pages =

    2022

  2. [2]

    2015 , publisher =

    Automatic Speech Recognition: A Deep Learning Approach , author =. 2015 , publisher =

    2015

  3. [3]

    2025 , type =

    Voice-based Interaction with Digital Services , author =. 2025 , type =

    2025

  4. [4]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume =

    Towards Building ASR Systems for the Next Billion Users , author =. Proceedings of the AAAI Conference on Artificial Intelligence , volume =. 2022 , doi =

    2022

  5. [5]

    Mon-Khmer Studies , volume =

    North-East India as a Linguistic Area , author =. Mon-Khmer Studies , volume =. 1997 , publisher =

    1997

  6. [6]

    The Sino-Tibetan Languages , editor =

    The Tibeto-Burman Languages of Northeastern India , author =. The Sino-Tibetan Languages , editor =. 2003 , publisher =

    2003

  7. [7]

    IEEE Access , pages =

    Towards low-resource languages machine translation: A language-specific fine-tuning with LoRA for specialized large language models , author=. IEEE Access , pages =. 2025 , publisher=

    2025

  8. [8]

    arXiv preprint arXiv:2307.07948 , year=

    Model adaptation for ASR in low-resource Indian languages , author=. arXiv preprint arXiv:2307.07948 , year=

    arXiv

Show all 136 references
  1. [9]

    Proceedings of the 6th Workshop on Spoken Language Technologies for Under-Resourced Languages (SLTU 2018) , pages =

    Development of Assamese Continuous Speech Recognition System , author =. Proceedings of the 6th Workshop on Spoken Language Technologies for Under-Resourced Languages (SLTU 2018) , pages =. 2018 , publisher =

    2018

  2. [10]

    Neural Networks , volume =

    Machine learning based sample extraction for automatic speech recognition using dialectal Assamese speech , author =. Neural Networks , volume =. 2016 , publisher =

    2016

  3. [11]

    2013 National Conference on Communications (NCC) , pages =

    Assamese spoken query system to access the price of agricultural commodities , author =. 2013 National Conference on Communications (NCC) , pages =. 2013 , publisher =

    2013

  4. [12]

    2024 Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) , pages =

    Data Selection using Spoken Language Identification for Low-Resource and Zero-Resource Speech Recognition , author =. 2024 Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) , pages =. 2024 , publisher =

    2024

  5. [13]

    2025 , doi =

    Chen, William and Tian, Jinchuan and Peng, Yifan and Yan, Brian and Yang, Chao-Han Huck and Watanabe, Shinji , journal =. 2025 , doi =

    2025

  6. [14]

    Indian Journal of Science and Technology , volume =

    Use of Bidirectional Long Short Term Memory in Spoken Word Detection with Reference to the Assamese Language , author =. Indian Journal of Science and Technology , volume =. 2022 , publisher =

    2022

  7. [15]

    International conference on machine learning , pages=

    Robust speech recognition via large-scale weak supervision , author=. International conference on machine learning , pages=. 2023 , organization=

    2023

  8. [16]

    arXiv preprint arXiv:2503.23542 , year =

    Whisper-LM: Improving ASR Models with Language Models for Low-Resource Languages , author =. arXiv preprint arXiv:2503.23542 , year =

    arXiv

  9. [17]

    EURASIP Journal on Audio, Speech, and Music Processing , volume =

    Exploration of Whisper fine-tuning strategies for low-resource ASR , author =. EURASIP Journal on Audio, Speech, and Music Processing , volume =. 2024 , publisher =

    2024

  10. [18]

    2004 , doi =

    Morris, Andrew Cameron and Maier, Viktoria and Green, Phil D , booktitle=. 2004 , doi =

    2004

  11. [19]

    Archives of Computational Methods in Engineering , volume =

    Automatic Speech Recognition System for Tonal Languages: State-of-the-Art Survey , author =. Archives of Computational Methods in Engineering , volume =. 2021 , publisher =

    2021

  12. [20]

    Encyclopedia of Language and Linguistics , edition =

    Speech Recognition, Automatic: History , author =. Encyclopedia of Language and Linguistics , edition =. 2006 , pages =

    2006

  13. [21]

    Development and Analysis of Speech Recognition Systems for Assamese Language Using

    Sarma, Himangshu and Saharia, Navanath and Sharma, Utpal , journal =. Development and Analysis of Speech Recognition Systems for Assamese Language Using. 2017 , publisher =

    2017

  14. [22]

    2015 International Symposium on Advanced Computing and Communication (ISACC) , pages =

    Isolated Assamese speech recognition using artificial neural network , author =. 2015 International Symposium on Advanced Computing and Communication (ISACC) , pages =. 2015 , publisher =

    2015

  15. [23]

    International Journal of Computing and Digital Systems , volume =

    Assamese Speech-Based Vocabulary Identification System Using Convolutional Neural Network , author =. International Journal of Computing and Digital Systems , volume =. 2022 , publisher =

    2022

  16. [24]

    , journal =

    Bharali, Sruti Sruba and Kalita, Sanjib Kr. , journal =. A Comparative Study of Different Features for Isolated Spoken Word Recognition Using. 2021 , publisher =

    2021

  17. [25]

    ACM Computing Surveys , volume =

    Survey of Hallucination in Natural Language Generation , author =. ACM Computing Surveys , volume =. 2023 , publisher =

    2023

  18. [26]

    arXiv preprint arXiv:2510.16567 , year =

    Hallucination Benchmark for Speech Foundation Models , author =. arXiv preprint arXiv:2510.16567 , year =

  19. [27]

    Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency (FAccT '24) , pages =

    Careless Whisper: Speech-to-Text Hallucination Harms , author =. Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency (FAccT '24) , pages =. 2024 , publisher =

    2024

  20. [28]

    and Davel, Marelie H

    De Vries, Nic J. and Davel, Marelie H. and Badenhorst, Jaco and Basson, Willem D. and De Wet, Febe and Barnard, Etienne and De Waal, Alta , journal =. A smartphone-based. 2014 , publisher =

    2014

  21. [29]

    2001 , url =

    Unicode Normalization Forms , author =. 2001 , url =

    2001

  22. [30]

    Electronics , volume =

    Implementation of a Whisper Architecture-Based Turkish Automatic Speech Recognition (ASR) System and Evaluation of the Effect of Fine-Tuning with a Low-Rank Adaptation (LoRA) Adapter on Its Performance , author =. Electronics , volume =. 2024 , publisher =

    2024

  23. [31]

    Neural Information Processing

    A Comprehensive Performance Evaluation of Whisper Models in Dysarthric Speech Recognition , author =. Neural Information Processing. ICONIP 2024 , series =. 2024 , publisher =

    2024

  24. [32]

    Time-domain algorithms for harmonic bandwidth reduction and time scaling of speech signals , year =

    Malah, David , journal =. Time-domain algorithms for harmonic bandwidth reduction and time scaling of speech signals , year =

  25. [33]

    Applied Sciences , volume =

    Analyzing Noise Robustness of Cochleogram and Mel Spectrogram Features in Deep Learning Based Speaker Recognition , author =. Applied Sciences , volume =. 2023 , publisher =

    2023

  26. [34]

    A Comparative Study of Deep Audio Models for Spectrogram-and Waveform-based SingFake Detection , year =

    Nguyen-Duc, Minh and Nguyen, Luong Vuong and Nguyen-Ho-Nhat, Huy and Nguyen, Tri-Hai and Lee, O-Joun , journal =. A Comparative Study of Deep Audio Models for Spectrogram-and Waveform-based SingFake Detection , year =

  27. [35]

    Journal of Industrial Information Integration , volume =

    Integrated end-to-end multilingual method for low-resource agglutinative languages using Cyrillic scripts , author =. Journal of Industrial Information Integration , volume =. 2025 , publisher =

    2025

  28. [36]

    International Journal of Electrical and Electronics , volume=

    Assamese numeral speech recognition using multiple features and cooperative LVQ-architectures , author=. International Journal of Electrical and Electronics , volume=

  29. [37]

    Oxford Research Encyclopedia of Linguistics , publisher =

    Languages of the World , author =. Oxford Research Encyclopedia of Linguistics , publisher =. 2018 , doi =

    2018

  30. [38]

    Journal of the International Phonetic Association , volume =

    Assamese , author =. Journal of the International Phonetic Association , volume =. 2012 , publisher =

    2012

  31. [39]

    2014 , doi =

    Phoneme-Based Speech Segmentation Using Hybrid Soft Computing Framework , author =. 2014 , doi =

    2014

  32. [40]

    Proceedings of the LFG 2019 Conference , pages =

    Assamese Case Alignment Shifts in Progress , author =. Proceedings of the LFG 2019 Conference , pages =. 2019 , address =

    2019

  33. [41]

    Journal of Advances in Information Technology , volume =

    Assamese Dialect Identification Using Static and Dynamic Features from Vowel , author =. Journal of Advances in Information Technology , volume =. 2024 , doi =

    2024

  34. [42]

    Georgia Institute of Technology

    Automatic speech recognition--a brief history of the technology development , author=. Georgia Institute of Technology. Atlanta Rutgers University and the University of California. Santa Barbara , volume=

  35. [43]

    The Journal of the Acoustical Society of America , volume =

    Automatic Recognition of Spoken Digits , author =. The Journal of the Acoustical Society of America , volume =. 1952 , publisher =

    1952

  36. [44]

    Readings in Speech Recognition , editor =

    Lowerre, Bruce , title =. Readings in Speech Recognition , editor =. 1990 , isbn =

    1990

  37. [45]

    Proceedings of the IEEE , volume =

    A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition , author =. Proceedings of the IEEE , volume =. 1989 , publisher =

    1989

  38. [46]

    The Journal of the Acoustical Society of America , volume=

    The harpy speech recognition system: performance with large vocabularies , author=. The Journal of the Acoustical Society of America , volume=. 1976 , publisher=

    1976

  39. [47]

    ICASSP '82

    Acoustic Pattern Matching and Beam Searching , author =. ICASSP '82. IEEE International Conference on Acoustics, Speech, and Signal Processing , volume =. 1982 , publisher =

    1982

  40. [48]

    Advances in Neural Information Processing Systems , volume =

    Wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations , author =. Advances in Neural Information Processing Systems , volume =. 2020 , doi =

    2020

  41. [49]

    Proceedings of the 35th Conference on Computational Linguistics and Speech Processing (ROCLING 2023) , pages=

    Accelerating Hakka Speech Recognition Research and Development Using the Whisper Model , author=. Proceedings of the 35th Conference on Computational Linguistics and Speech Processing (ROCLING 2023) , pages=

    2023

  42. [50]

    IEEE Access , volume=

    Enhancing automatic speech recognition: effects of semantic audio filtering on models performance , author=. IEEE Access , volume=. 2024 , publisher=

    2024

  43. [51]

    Bhandari, Abhishek and Harit, Gaurav , journal =. Post-. 2026 , publisher =

    2026

  44. [53]

    arXiv preprint arXiv:2412.15726 , year =

    Fine-tuning Whisper on Low-Resource Languages for Real-World Applications , author =. arXiv preprint arXiv:2412.15726 , year =

    arXiv

  45. [54]

    Proceedings of the Sixth Workshop on African Natural Language Processing (AfricaNLP 2025) , pages =

    Fine-tuning Whisper Tiny for Swahili ASR: Challenges and Recommendations for Low-Resource Speech Recognition , author =. Proceedings of the Sixth Workshop on African Natural Language Processing (AfricaNLP 2025) , pages =. 2025 , publisher =

    2025

  46. [55]

    Proceedings of the 12th Language Resources and Evaluation Conference (LREC 2020) , pages =

    Common Voice: A Massively-Multilingual Speech Corpus , author =. Proceedings of the 12th Language Resources and Evaluation Conference (LREC 2020) , pages =. 2020 , publisher =

    2020

  47. [56]

    Proceedings of the 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages =

    Canonicalization of Database Records Using Adaptive Similarity Measures , author =. Proceedings of the 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages =. 2007 , publisher =

    2007

  48. [57]

    Computers in Biology and Medicine , volume =

    Two-stage Data Augmentation for Improved ASR Performance for Dysarthric Speech , author =. Computers in Biology and Medicine , volume =. 2025 , publisher =

    2025

  49. [58]

    Common Voice Dataset , year =

  50. [59]

    arXiv preprint arXiv:1706.02677 , year=

    Accurate, large minibatch sgd: Training imagenet in 1 hour , author=. arXiv preprint arXiv:1706.02677 , year=

    arXiv

  51. [60]

    Advances in Neural Information Processing Systems 4 , pages =

    A Simple Weight Decay Can Improve Generalization , author =. Advances in Neural Information Processing Systems 4 , pages =. 1991 , publisher =

    1991

  52. [61]

    Advances in Neural Information Processing Systems 1 , pages =

    Comparing Biases for Minimal Network Construction with Back-Propagation , author =. Advances in Neural Information Processing Systems 1 , pages =. 1988 , publisher =

    1988

  53. [62]

    Proceedings of the 7th International Conference on Learning Representations (ICLR) , year =

    Decoupled Weight Decay Regularization , author =. Proceedings of the 7th International Conference on Learning Representations (ICLR) , year =

  54. [64]

    Applied Computational Intelligence and Soft Computing , volume =

    Segmentation and Classification of Vowel Phonemes of Assamese Speech Using a Hybrid Neural Framework , author =. Applied Computational Intelligence and Soft Computing , volume =. 2012 , publisher =

    2012

  55. [65]

    Frontiers in Human Neuroscience , volume =

    Intersections between Heritage, Multilingualism, and Education: Language Acquisition in India , author =. Frontiers in Human Neuroscience , volume =. 2025 , publisher =

    2025

  56. [66]

    Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics , pages =

    BLEU: A Method for Automatic Evaluation of Machine Translation , author =. Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics , pages =. 2002 , address =

    2002

  57. [67]

    2005 , publisher =

    Banerjee, Satanjeev and Lavie, Alon , booktitle =. 2005 , publisher =

    2005

  58. [69]

    Journal of Big Data , volume =

    Evaluation of the effectiveness and efficiency of state-of-the-art features and models for automatic speech recognition error detection , author =. Journal of Big Data , volume =. 2021 , publisher =

    2021

  59. [70]

    arXiv preprint arXiv:2510.07221 , year =

    How Much Speech Data Is Necessary for ASR in African Languages? An Evaluation of Data Scaling in Kinyarwanda and Kikuyu , author =. arXiv preprint arXiv:2510.07221 , year =

  60. [71]

    Speech Communication , volume =

    A Smartphone-Based ASR Data Collection Tool for Under‑Resourced Languages , author =. Speech Communication , volume =. 2014 , publisher =

    2014

  61. [72]

    Proceedings of the 2024 International Conference on Decision Aid Sciences and Applications (DASA) , pages =

    Assessing Speech-to-Text Translation Quality: An Overview of Key Metrics , author =. Proceedings of the 2024 International Conference on Decision Aid Sciences and Applications (DASA) , pages =. 2024 , publisher =

    2024

  62. [73]

    , author Sarma, K.K

    author Agarwalla, S. , author Sarma, K.K. , year 2016 . title Machine learning based sample extraction for automatic speech recognition using dialectal assamese speech . journal Neural Networks volume 78 , pages 97--111 . :10.1016/j.neunet.2015.12.010

    2016 doi

  63. [74]

    , author Nafula, E

    author Akera, B. , author Nafula, E. , author Walukagga, P. , author Yiga, G. , author Quinn, J. , author Mwebaze, E. , year 2025 . title How much speech data is necessary for asr in african languages? an evaluation of data scaling in kinyarwanda and kikuyu . journal arXiv pre...

    2025 doi

  64. [75]

    , author Branson, M

    author Ardila, R. , author Branson, M. , author Davis, K. , author Kohler, M. , author Meyer, J. , author Henretty, M. , author Morais, R. , author Saunders, L. , author Tyers, F. , author Weber, G. , year 2020 . title Common voice: A massively-multilingual speech corpus , in:...

    2020 doi arXiv

  65. [76]

    , author Zhou, Y

    author Baevski, A. , author Zhou, Y. , author Mohamed, A. , author Auli, M. , year 2020 . title Wav2vec 2.0: A framework for self-supervised learning of speech representations . journal Advances in Neural Information Processing Systems volume 33 , pages 12449--12460 . :10.4855...

    2020 doi arXiv

  66. [77]

    , author Lavie, A

    author Banerjee, S. , author Lavie, A. , year 2005 . title METEOR : An automatic metric for MT evaluation with improved correlation with human judgments , in: booktitle Proceedings of the ACL Workshop on Intrinsic and Extrinsic Evaluation Measures for Machine Translation and/o...

    2005

  67. [78]

    , author Razaque, A

    author Bekarystankyzy, A. , author Razaque, A. , author Mamyrbayev, O. , year 2025 . title Integrated end-to-end multilingual method for low-resource agglutinative languages using cyrillic scripts . journal Journal of Industrial Information Integration volume 43 , pages 100750...

    2025

  68. [79]

    , author Harit, G

    author Bhandari, A. , author Harit, G. , year 2026 . title Post- ASR correction for low-resource Rajasthani language . journal ACM Transactions on Asian and Low-Resource Language Information Processing :10.1145/3793254

    2026 doi

  69. [80]

    , author Kalita, S.K

    author Bharali, S.S. , author Kalita, S.K. , year 2021 . title A comparative study of different features for isolated spoken word recognition using HMM with reference to assamese language . journal International Journal of Speech Technology volume 18 , pages 673--684 . :10.100...

    2021 doi

  70. [81]

    , author Strik, H

    author Bhat, C. , author Strik, H. , year 2025 . title Two-stage data augmentation for improved asr performance for dysarthric speech . journal Computers in Biology and Medicine volume 189 , pages 109954 . :10.1016/j.compbiomed.2025.109954

    2025

  71. [82]

    , year 2003

    author Burling, R. , year 2003 . title The tibeto-burman languages of northeastern india , in: editor Thurgood, G. , editor LaPolla, R.J. (Eds.), booktitle The Sino-Tibetan Languages . edition 1st ed.. publisher Routledge , address London and New York . Routledge Language Fami...

    2003 doi

  72. [83]

    , author Hsu, Y.H

    author Chen, C.Y. , author Hsu, Y.H. , author Chang, C.c. , year 2023 . title Accelerating hakka speech recognition research and development using the whisper model , in: booktitle Proceedings of the 35th Conference on Computational Linguistics and Speech Processing (ROCLING 2...

    2023

  73. [84]

    , author Chu, C

    author Chen, J. , author Chu, C. , author Li, S. , author Kawahara, T. , year 2024 . title Data selection using spoken language identification for low-resource and zero-resource speech recognition , in: booktitle 2024 Asia Pacific Signal and Information Processing Association ...

    2024

  74. [85]

    , author Tian, J

    author Chen, W. , author Tian, J. , author Peng, Y. , author Yan, B. , author Yang, C.H.H. , author Watanabe, S. , year 2025 . title OWLS : Scaling laws for multilingual speech recognition and translation models . journal arXiv preprint arXiv:2502.10373 :10.48550/arXiv.2502.10373

    2025 doi arXiv

  75. [86]

    , author Wick, M.L

    author Culotta, A. , author Wick, M.L. , author Hall, R.J. , author Marzilli, M. , author McCallum, A. , year 2007 . title Canonicalization of database records using adaptive similarity measures , in: booktitle Proceedings of the 13th ACM SIGKDD International Conference on Kno...

    2007

  76. [87]

    , author Bhattacharjee, U

    author Das, H.C. , author Bhattacharjee, U. , year 2024 . title Assamese dialect identification using static and dynamic features from vowel . journal Journal of Advances in Information Technology volume 15 , pages 306--317 . :10.12720/jait.15.2.306-321

    2024 doi

  77. [88]

    , author Biddulph, R

    author Davis, K.H. , author Biddulph, R. , author Balashek, S. , year 1952 . title Automatic recognition of spoken digits . journal The Journal of the Acoustical Society of America volume 24 , pages 637--642 . :10.1121/1.1906946

    1952 doi

  78. [89]

    , author D \"u rst, M

    author Davis, M. , author D \"u rst, M. , year 2001 . title Unicode Normalization Forms . type Unicode Standard Annex \#15 . Unicode Consortium. https://www.unicode.org/reports/tr15/

    2001

  79. [90]

    , author Davel, M.H

    author De Vries, N.J. , author Davel, M.H. , author Badenhorst, J. , author Basson, W.D. , author De Wet, F. , author Barnard, E. , author De Waal, A. , year 2014 . title A smartphone-based asr data collection tool for under‑resourced languages . journal Speech Communication v...

    2014 doi

  80. [91]

    , author Nirmala, S.R

    author Deka, B. , author Nirmala, S.R. , author Samudravijaya, K. , year 2018 . title Development of assamese continuous speech recognition system , in: booktitle Proceedings of the 6th Workshop on Spoken Language Technologies for Under-Resourced Languages (SLTU 2018) , publis...

    2018 doi

  81. [92]

    , author Choudhury, R.D

    author Dutta, D. , author Choudhury, R.D. , author Barman, U. , year 2022 . title Assamese speech-based vocabulary identification system using convolutional neural network . journal International Journal of Computing and Digital Systems volume 12 , pages 1191--1202 . :10.12785...

    2022 doi

  82. [93]

    , author Errattahi, R

    author El Hannani, A. , author Errattahi, R. , author Salmam, F.Z. , author Hain, T. , author Ouahmane, H. , year 2021 . title Evaluation of the effectiveness and efficiency of state-of-the-art features and models for automatic speech recognition error detection . journal Jour...

    2021 doi

  83. [94]

    , author Doll \'a r, P

    author Goyal, P. , author Doll \'a r, P. , author Girshick, R. , author Noordhuis, P. , author Wesolowski, L. , author Kyrola, A. , author Tulloch, A. , author Jia, Y. , author He, K. , year 2017 . title Accurate, large minibatch sgd: Training imagenet in 1 hour . journal arXi...

    2017 doi arXiv

  84. [95]

    , author Lowerre, B

    author Greer, K. , author Lowerre, B. , author Wilcox, L. , year 1982 . title Acoustic pattern matching and beam searching , in: booktitle ICASSP '82. IEEE International Conference on Acoustics, Speech, and Signal Processing , publisher IEEE . pp. pages 1251--1254 . :10.1109/I...

    1982

  85. [96]

    , author Pratt, L.Y

    author Hanson, S.J. , author Pratt, L.Y. , year 1988 . title Comparing biases for minimal network construction with back-propagation , in: booktitle Advances in Neural Information Processing Systems 1 , publisher Morgan Kaufmann . pp. pages 177--185

    1988

  86. [97]

    , author Doddapaneni, S

    author Javed, T. , author Doddapaneni, S. , author Raman, A. , author Bhogale, K.S. , author Ramesh, G. , author Kunchukuttan, A. , author Kumar, P. , author Khapra, M.M. , year 2022 . title Towards building asr systems for the next billion users , in: booktitle Proceedings of...

    2022 doi

  87. [98]

    , author Lee, N

    author Ji, Z. , author Lee, N. , author Frieske, R. , author Yu, T. , author Su, D. , author Xu, Y. , author Ishii, E. , author Bang, Y.J. , author Madotto, A. , author Fung, P. , year 2023 . title Survey of hallucination in natural language generation . journal ACM Computing ...

    2023 doi

  88. [99]

    , author Borbora, K.A

    author Kalita, D. , author Borbora, K.A. , author Nath, D. , year 2022 . title Use of bidirectional long short term memory in spoken word detection with reference to the assamese language . journal Indian Journal of Science and Technology volume 15 , pages 1364--1371 . :10.174...

    2022 doi

  89. [100]

    , author Dr z \' k, D

    author Kapusta, J. , author Dr z \' k, D. , author S teflovi c , K. , author Nagy, K.S. , year 2024 . title Text data augmentation techniques for word embeddings in fake news classification . journal IEEE Access volume 12 , pages 31538--31550 . :10.1109/ACCESS.2024.3369918

    2024

  90. [101]

    , author Singh, A

    author Kaur, J. , author Singh, A. , author Kadyan, V. , year 2021 . title Automatic speech recognition system for tonal languages: State-of-the-art survey . journal Archives of Computational Methods in Engineering volume 28 , pages 1039--1068 . :10.1007/s11831-020-09414-4

    2021 doi

  91. [102]

    , author Choi, A.S.G

    author Koenecke, A. , author Choi, A.S.G. , author Mei, K.X. , author Schellmann, H. , author Sloane, M. , year 2024 . title Careless whisper: Speech-to-text hallucination harms , in: booktitle Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparenc...

    2024

  92. [103]

    , author La Quatra, M

    author Koudounas, A. , author La Quatra, M. , author Giollo, M. , author Siniscalchi, S.M. , author Baralis, E. , year 2025 . title Hallucination benchmark for speech foundation models . journal arXiv preprint arXiv:2510.16567 :10.48550/arXiv.2510.16567

    2025 doi

  93. [104]

    , author Belangour, A

    author Labied, M. , author Belangour, A. , author Banane, M. , year 2024 . title Assessing speech-to-text translation quality: An overview of key metrics , in: booktitle Proceedings of the 2024 International Conference on Decision Aid Sciences and Applications (DASA) , publish...

    2024

  94. [105]

    , author Srinivasagan, R

    author Lambamo, W. , author Srinivasagan, R. , author Jifara, W. , year 2023 . title Analyzing noise robustness of cochleogram and mel spectrogram features in deep learning based speaker recognition . journal Applied Sciences volume 13 , pages 569 . :10.3390/app13010569

    2023 doi

  95. [106]

    , year 2018

    author Leben, W.R. , year 2018 . title Languages of the world , in: booktitle Oxford Research Encyclopedia of Linguistics . publisher Oxford University Press . :10.1093/acrefore/9780199384655.013.349

    2018

  96. [107]

    , author Khaw, Y.M.J

    author Liang, X. , author Khaw, Y.M.J. , author Liew, S.Y. , author Tan, T.P. , author Qin, D. , year 2025 . title Towards low-resource languages machine translation: A language-specific fine-tuning with lora for specialized large language models . journal IEEE Access , pages ...

    2025

  97. [108]

    , author Yang, X

    author Liu, Y. , author Yang, X. , author Qu, D. , year 2024 . title Exploration of whisper fine-tuning strategies for low-resource asr . journal EURASIP Journal on Audio, Speech, and Music Processing volume 2024 , pages 29 . :10.1186/s13636-024-00349-3

    2024 doi

  98. [109]

    , author Hutter, F

    author Loshchilov, I. , author Hutter, F. , year 2019 . title Decoupled weight decay regularization , in: booktitle Proceedings of the 7th International Conference on Learning Representations (ICLR) , address New Orleans, LA, USA . :10.48550/arXiv.1711.05101

    2019 doi arXiv

  99. [110]

    , year 1990

    author Lowerre, B. , year 1990 . title The HARPY speech understanding system , in: editor Waibel, A. , editor Lee, K.F. (Eds.), booktitle Readings in Speech Recognition . publisher Morgan Kaufmann Publishers Inc. , address San Francisco, CA, USA , pp. pages 576--586 . :10.1016...

    1990 doi

  100. [111]

    , author Vasconcelos, D.J.R

    author Magalhães, R.P. , author Vasconcelos, D.J.R. , author Fernandes, G.S. , author Cruz, L.A. , author Sampaio, M.X. , author de Macêdo, J.A.F. , author da Silva, T.L.C. , year 2022 . title Evaluation of automatic speech recognition approaches . journal Journal of Informati...

    2022

  101. [112]

    , year 2012

    author Mahanta, S. , year 2012 . title Assamese . journal Journal of the International Phonetic Association volume 42 , pages 217--224 . :10.1017/S0025100312000096

    2012 doi

  102. [113]

    , year 1979

    author Malah, D. , year 1979 . title Time-domain algorithms for harmonic bandwidth reduction and time scaling of speech signals . journal IEEE Transactions on Acoustics, Speech, and Signal Processing volume 27 , pages 121--133 . :10.1109/TASSP.1979.1163210

    1979

  103. [114]

    , author Talukdar, P.H

    author Medhi, B. , author Talukdar, P.H. , year 2015 . title Isolated assamese speech recognition using artificial neural network , in: booktitle 2015 International Symposium on Advanced Computing and Communication (ISACC) , publisher IEEE . pp. pages 141--148 . :10.1109/ISACC...

    2015

  104. [115]

    , year 1997

    author Moral, D. , year 1997 . title North-east india as a linguistic area . journal Mon-Khmer Studies volume 27 , pages 43--54

    1997

  105. [116]

    , author Maier, V

    author Morris, A.C. , author Maier, V. , author Green, P.D. , year 2004 . title From WER and RIL to MER and WIL: improved evaluation measures for connected speech recognition. , in: booktitle Interspeech , pp. pages 2765--2768 . :doi: 10.21437/Interspeech.2004-668

    2004 doi

  106. [117]

    title Common voice dataset

    author Mozilla Foundation , year 2026 . title Common voice dataset . howpublished https://datacollective.mozillafoundation.org/ . note Accessed: 7 Mar 2026

    2026

  107. [118]

    , author Nguyen, L.V

    author Nguyen-Duc, M. , author Nguyen, L.V. , author Nguyen-Ho-Nhat, H. , author Nguyen, T.H. , author Lee, O.J. , year 2025 . title A comparative study of deep audio models for spectrogram-and waveform-based singfake detection . journal IEEE Access volume 13 , pages 15312--15...

    2025

  108. [119]

    , author Roukos, S

    author Papineni, K. , author Roukos, S. , author Ward, T. , author Zhu, W.J. , year 2002 . title Bleu: A method for automatic evaluation of machine translation , in: booktitle Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics , publisher A...

    2002

  109. [120]

    , author Santos, T

    author Perezhohin, Y. , author Santos, T. , author Costa, V. , author Peres, F. , author Castelli, M. , year 2024 . title Enhancing automatic speech recognition: effects of semantic audio filtering on models performance . journal IEEE Access volume 12 , pages 155136--155150 . ...

    2024

  110. [121]

    , author Turan, A.K

    author Polat, H. , author Turan, A.K. , author Ko c ak, C. , author Ula s , H.B. , year 2024 . title Implementation of a whisper architecture-based turkish automatic speech recognition (asr) system and evaluation of the effect of fine-tuning with a low-rank adaptation (lora) a...

    2024 doi

  111. [122]

    , author Rao, V.G.S

    author Prasad, C. , author Rao, V.G.S. , author Naidu, R.C.A. , et al., year 2026 . title An asr transformer-based model for kannada speech-to-text transcription . journal Journal of Artificial Intelligence and Technology :10.37965/jait.2026.0935

    2026

  112. [123]

    , year 1989

    author Rabiner, L.R. , year 1989 . title A tutorial on hidden markov models and selected applications in speech recognition . journal Proceedings of the IEEE volume 77 , pages 257--286 . :10.1109/5.18626

    1989 doi

  113. [124]

    , author Juang, B.H

    author Rabiner, L.R. , author Juang, B.H. , year 2006 . title Speech recognition, automatic: History , in: editor Brown, K. (Ed.), booktitle Encyclopedia of Language and Linguistics . edition 2 ed.. publisher Elsevier , pp. pages 806--819 . :10.1016/B0-08-044854-2/00906-8

    2006 doi

  114. [125]

    , author Kim, J.W

    author Radford, A. , author Kim, J.W. , author Xu, T. , author Brockman, G. , author McLeavey, C. , author Sutskever, I. , year 2023 . title Robust speech recognition via large-scale weak supervision , in: booktitle International conference on machine learning , organization P...

    2023 doi arXiv

  115. [126]

    , author Camilleri, M

    author Saikia, P. , author Camilleri, M. , year 2019 . title Assamese case alignment shifts in progress , in: booktitle Proceedings of the LFG 2019 Conference , publisher CSLI Publications , address Canberra, Australia . pp. pages 251--271

    2019

  116. [127]

    , author Saharia, N

    author Sarma, H. , author Saharia, N. , author Sharma, U. , year 2017 . title Development and analysis of speech recognition systems for assamese language using HTK . journal ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP) volume 17 , pages ...

    2017 doi

  117. [128]

    , author Sarma, K.K

    author Sarma, M. , author Sarma, K.K. , year 2012 . title Segmentation and classification of vowel phonemes of assamese speech using a hybrid neural framework . journal Applied Computational Intelligence and Soft Computing volume 2012 , pages 1--8 . :10.1155/2012/871324

    2012 doi

  118. [129]

    , author Sarma, K.K

    author Sarma, M. , author Sarma, K.K. , year 2014 . title Phoneme-Based Speech Segmentation Using Hybrid Soft Computing Framework . volume volume 550 of series Studies in Computational Intelligence . publisher Springer International Publishing , address Cham, Switzerland . :10...

    2014 doi

  119. [130]

    , author Sarma, K.K

    author Sarma, M.P. , author Sarma, K.K. , year 2011 . title Assamese numeral speech recognition using multiple features and cooperative lvq-architectures . journal International Journal of Electrical and Electronics volume 5 , pages 1

    2011

  120. [131]

    , year 2025

    author Sarma, V.M. , year 2025 . title Intersections between heritage, multilingualism, and education: Language acquisition in india . journal Frontiers in Human Neuroscience volume 19 , pages 1538482 . :10.3389/fnhum.2025.1538482

    2025

  121. [132]

    , author Thotappa, D

    author Shahnawazuddin, S. , author Thotappa, D. , author Sarma, B.D. , author Deka, A. , author Prasanna, S.R.M. , author Sinha, R. , year 2013 . title Assamese spoken query system to access the price of agricultural commodities , in: booktitle 2013 National Conference on Comm...

    2013

  122. [133]

    , author Pandya, M

    author Sharma, A.K. , author Pandya, M. , author Shukla, A. , year 2025 . title Fine-tuning whisper tiny for swahili asr: Challenges and recommendations for low-resource speech recognition , in: booktitle Proceedings of the Sixth Workshop on African Natural Language Processing...

    2025 doi

  123. [134]

    , author Mehta, A.S

    author Singh, A. , author Mehta, A.S. , author Nanavati, J. , author Bandekar, J. , author Basumatary, K. , author Badiger, S. , author Udupa, S. , author Kumar, S. , author Ghosh, P.K. , author Pai, P. , et al., year 2023 . title Model adaptation for asr in low-resource india...

    2023 doi arXiv

  124. [135]

    , author Zhong, Z

    author Singh, S. , author Zhong, Z. , author Wang, Q. , author Mendes, C. , author Hasegawa-Johnson, M. , author Abdulla, W. , author Shahamiri, S.R. , year 2024 . title A comprehensive performance evaluation of whisper models in dysarthric speech recognition , in: booktitle N...

    2024 doi

  125. [136]

    , author Javaid, A

    author Sourav, M.S.G. , author Javaid, A. , author Cheng, L. , year 2025 . title A review of ai in human‑machine cooperation: Machine perspective . journal ACM Transactions on Autonomous and Adaptive Systems volume 20 , pages 1--24 . :10.1145/3774318

    2025 doi

  126. [137]

    , author Paonessa, C

    author Timmel, V. , author Paonessa, C. , author Vogel, M. , author Perruchoud, D. , author Kakooee, R. , year 2024 . title Fine-tuning whisper on low-resource languages for real-world applications . journal arXiv preprint arXiv:2412.15726 :10.48550/arXiv.2412.15726

    2024 doi arXiv

  127. [138]

    , author Deng, L

    author Yu, D. , author Deng, L. , year 2015 . title Automatic Speech Recognition: A Deep Learning Approach . volume volume 1 of series Signals and Communication Technology . publisher Springer , address London . :10.1007/978-1-4471-5779-3

    2015 doi

  128. [139]

    , author Navas, E

    author de Zuazo, X. , author Navas, E. , author Saratxaga, I. , author Rioja, I.H. , year 2025 . title Whisper-lm: Improving asr models with language models for low-resource languages . journal arXiv preprint arXiv:2503.23542 :10.48550/arXiv.2503.23542

    2025 doi arXiv

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.