A Machine Hearing System for Robust Cough Detection Based on a High-Level Representation of Band-Specific Audio Features

Carlos Hoyos-Barcel\'o; Jes\'us Monge-Alvarez; Luis M. San-Jos\'e-Revuelta; Pablo Casaseca-de-la-Higuera

REVIEW 4 major objections 5 minor 43 references

A Machine Hearing System for Robust Cough Detection Based on a High-Level Representation of Band-Specific Audio Features

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Simple audio features beat deep learning on cough detection

desk verdict A coherent applied cough-detection paper with a genuine method contribution, but the evaluation has a feature-selection leakage that undercuts the headline comparison until fixed. read the letter →

arxiv 2412.01996 v1 pith:DHAWEBY5 submitted 2024-12-02 eess.AS cs.SDeess.SP

Jesús Monge-Alvarez , Carlos Hoyos-Barceló , Luis M. San-José-Revuelta , Pablo Casaseca-de-la-Higuera This is my paper · ORCID

classification eess.AScs.SDeess.SP

keywords coughdetectionmachinehearingaudioclassificationspectralfeaturesfrequencybandssupportvectorfeatureselectionnoiserobustness

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

This paper claims that a lightweight machine hearing system can detect coughs in noisy real-life audio more reliably than previously proposed methods, including a convolutional neural network. The system computes a battery of spectral features separately in five frequency bands, selects 29 of them that stay informative across three noise conditions, and summarizes short-term features by mean and standard deviation over 300 ms windows before feeding an SVM. On recordings from thirteen patients in three simulated real-life scenarios, it reports 92.71% sensitivity, 88.58% specificity, and a 90.69% area under the ROC curve. If correct, the finding matters because it suggests accurate, real-time cough monitoring can run on a smartphone without the computational cost of deep learning.

What carries the argument

The machinery is the combination of five predefined frequency bands with a mean/standard-deviation summary (AvgSD) over long-term frames. Each band [0,0.5), [0.5,1), [1,1.5), [1.5,2), [2,5.5125] kHz acts as a separate channel in which spectral centroid, bandwidth, crest factor, flatness, flux, roll-off, f50/f90 ratio, peak entropy, Renyi entropy, kurtosis, skewness, and relative power are computed from a Welch PSD; computing these descriptors per band localizes the impact of colored ambient noise to the affected bands. The 300 ms long-term window matches the average cough duration and, by concatenating feature-wise means and standard deviations, compresses the short-term stream into a compact vector that an SVM can separate robustly.

What would settle it

Re-annotate the same thirteen hours of recordings with at least one additional independent annotator using the same frame-based rule, then measure inter-annotator agreement; if agreement is low, or if retraining and testing the pipeline on each annotator's labels yields substantially different sensitivity/specificity from the reported 92.71% and 88.58%, the central claim is not settled.

Watch

Extended reading notes

Core claim

The central claim is that cough detection in noisy environments is better served by band-specific short-term spectral features with a high-level mean-and-standard-deviation representation than by the end-to-end deep networks and global-spectrum feature sets used in prior work. The authors construct 117 short-term features—twelve spectral descriptors computed independently in each of five bands covering [0, 5.5125] kHz, plus spectral entropy and standard audio descriptors—then reduce them by Relieff selection combined across three protocol parts to a 29-dimensional set. Long-term frames of about 300 ms are formed from five overlapping 75 ms frames, and the per-feature mean and standard deviation constitute the high-level representation that feeds a polynomial-kernel SVM. In block-wise cross-validation, this AvgSD pipeline yields the reported figures in all three noise scenarios and outperforms the ensemble-subband method of You et al., the CNN of Amoh and Odame, and a moment-based cepstrogram method, with McNemar's test indicating significance for sensitivity and specificity.

Load-bearing premise

The evaluation rests on the assumption that the manually annotated frame labels in the thirteen-patient database are correct and consistent; if the labels are noisy or biased, both training and evaluation would be affected, and the reported sensitivity and specificity could be misleading.

Editorial extensions

If this is right

A smartphone-based cough monitor could run continuously in real time because the final feature set has only 29 dimensions and the classifier is a simple SVM with a polynomial kernel.
The band-specific feature computation appears to make the detector robust across noise levels from low (speech/reading) to high (patient movement plus TV or radio), so a single model can be deployed without first classifying the acoustic environment.
Leave-one-patient-out experiments suggest the system generalizes to new patients in terms of sensitivity and specificity mostly above 80%, with one patient (ID 6) consistently lower.
The comparison results imply that a handcrafted feature pipeline can outperform a CNN for cough detection when the positive class is small and the negative class is diverse, which may guide architecture choices in other audio event detection tasks.
Training three noise-specific models and combining them by majority voting increases specificity relative to a single model, at the cost of some sensitivity; both configurations keep sensitivity in the 90% range.

Reading between the lines

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

Inference: the same band-specific plus mean/std recipe could transfer to other impulsive bioacoustic events with sparse spectra, such as sneezes, snoring, or throat clearing, by re-running the Relieff selection on the new target labels.
Inference: because the paper shows the selected features are largely robust across its three noise parts, one testable extension is to evaluate the fixed 29-feature set on additive noise at SNRs below -6 dB or on unseen recording devices, where the band-localization argument predicts graceful degradation.
Inference: the authors do not quantify how much of the gain comes from the band decomposition versus the AvgSD representation; an ablation that computes the same features over the full spectrum with and without mean/std, or over the five bands with short-term classification, would isolate the contribution of each design choice.
Inference: the clinical rationale implies that secondary measures—cough rate, cough intensity, dry versus wet cough—could be derived from the same high-level features, but the paper does not demonstrate this; a follow-up study could test whether the 29 features separate cough subtypes.

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

What should you know: this is a solid engineering paper from 2019, now on arXiv, with a plausible method and a real evaluation weakness that matters for the headline claim. The band-specific computation of 12 spectral features plus mean/std over 300 ms frames is a genuine combination not present in the cited prior work, and the authors compare fairly against three recent detectors on a 13-patient database covering three noise conditions. If the numbers are trustworthy, the system beats a CNN and two handcrafted-feature baselines, which is a useful result for phone-based cough monitoring.

Credit where due: the pipeline is clearly described; the selection process is detailed enough to spot the issue; the leave-one-patient-out experiments are the right generalization check; and McNemar tests on the main comparisons are a step up from just reporting means. The paper also acknowledges the small population and notes that the state-of-the-art methods were originally evaluated on easier train/test splits.

Now the soft spot, and it is a real one. Section III-B runs Relieff on 10% of observations per protocol part, and IV-A repeats this five times on disjoint subsets covering 50% of observations, then picks 29 features. The only safeguard stated is that no long-term features from the train or test sets of the Fig. 7 folds were used for selection. But Relieff runs on short-term frames, and the statement does not say those short-term frames came only from training blocks or training patients. If any selected short-term frame belongs to a block or patient later used for testing, the feature ranking is supervised by test labels and the Table IV/V comparison is biased in favor of the proposed system. The stress-test note is correct on this point. The fix is standard: nest feature selection inside each training fold and report per-fold and per-patient intervals. Also, the abstract quotes only the Part 1 numbers (92.71 SEN, 88.58 SPE); the final system's specificity is closer to 80%, so the abstract is misleading. Minor: no inter-annotator reliability for the frame labels, and no code/data, so the numbers cannot be independently checked.

Bottom line: if the leakage is real, the quantitative edge over prior work is not established, but the method itself is coherent and the band-specific features could still be valuable. The paper is worth a serious referee: an editor should send it to review, and the review should push for nested feature selection and honest headline numbers. I'd bring it to a reading group as an example of a clearly written applied ML paper with an evaluation pitfall.

Referee Report

4 major / 5 minor

Summary. The paper proposes a two-stage machine-hearing system for cough detection from audio. Short-term spectral features are computed in five predefined frequency bands and combined with standard audio descriptors; a supervised Relieff-based procedure selects 29 features; a high-level representation is obtained either by mean/standard deviation (AvgSD) of 300-ms long-term frames or by supervised bag-of-audio-words; classification is performed by a polynomial-kernel SVM. The system is evaluated on 13 patients recorded under three noise scenarios, with per-part models, a leave-one-patient-out cross-validation, and a final ensemble. The abstract reports 92.71% sensitivity, 88.58% specificity, and 90.69% AUC for the first (low-noise) part, and the paper claims the approach outperforms state-of-the-art methods including a CNN.

Significance. If the reported results are valid, the paper would demonstrate that a lightweight handcrafted-feature pipeline with a high-level AvgSD representation can match or beat deep-learning cough detectors on noisy real-life recordings, which is practically relevant for smartphone-based continuous monitoring. The study has clear strengths: evaluation in three protocol-defined noise environments, leave-one-patient-out cross-validation, statistical comparison via McNemar's test, and an explicit acknowledgment of the small dataset as a limitation. However, the central claim of superiority is currently supported only by the per-part models, and the evaluation protocol has a potentially serious feature-selection leakage problem that directly affects the headline numbers. The significance of the contribution is therefore conditional on the evaluation being corrected.

major comments (4)

[III-B, IV-A, III-D] The feature selection procedure appears to use test labels before the cross-validation splits are formed. Section III-B selects 10% of observations per protocol part with Relieff, and Section IV-A repeats this five times on disjoint subsets, totaling 50% of the observations. The block-wise five-fold partitions described in Section III-D are then applied afterward. The only safeguard stated at the end of Section IV-A is that 'no long-term features from the train or test sets in each of the 5 folds in Fig. 7 was employed for selection,' which addresses long-term features only, not the short-term frames to which Relieff was applied. Since Relieff is supervised, any randomly selected short-term frame belonging to a later test block or test patient can influence the chosen 29 features, inflating the results in Table IV and Table V and biasing the comparison with [15], [17], and [18]. The authors must either prove that no observation used in selection ever appears in a test fold, or rerun the whole pipeline with feature selection nested inside each training fold and report the resulting SEN/SPE/AUC.
[Abstract, Table IV, Fig. 10, Conclusions] The abstract's headline numbers (92.71% sensitivity, 88.58% specificity, 90.69% AUC) are taken from the first part of the protocol in Table IV, which is the low-noise scenario, not from the final deployed system. The final system, whether the single model in Fig. 10(b) or the majority-voting ensemble in Fig. 10(a), achieves sensitivity in the 90% range but specificity around 80%. The conclusion that the system 'outperform[s] so far proposed methods in terms of cough detection' is therefore not supported by the final-system evaluation. The paper should either report the final-system comparison against the state-of-the-art methods, or clearly qualify the headline claim as applying to the per-part models and not to the system as a whole.
[II] The manual annotation procedure is described only as 'Audio files were manually annotated on a time-frame basis' with majority-class assignment for mixed frames. No inter-annotator reliability, annotation protocol details, or confidence measures are reported. Since these labels define both the training targets and the ground truth for evaluation, label noise or systematic bias could directly affect the reported sensitivity and specificity, regardless of classifier performance. The authors should provide inter-annotator agreement statistics or, at minimum, a sensitivity analysis showing that the main conclusions are robust to plausible annotation errors.
[IV-B, Table IV, Table V, Fig. 8] The abstract and conclusions state an AUC value of 90.69%, but Tables IV and V report only SEN, SPE, and ACC, and no numeric AUC values are given for any model or comparison. ROC curves are shown in Fig. 8, but the AUC value in the abstract cannot be traced to a specific table entry. The authors should report AUC values (with confidence intervals) for each part and for the state-of-the-art methods, and state explicitly which model the abstract's AUC refers to.

minor comments (5)

[III-D] The phrase '75 ms frames with 19s overlap' in Section III-D appears to contain a typo: the overlap should be 19 ms, as confirmed by the calculation (75−19)·4+75 = 299 ms.
[III-A1] The fifth frequency band is defined inconsistently as [2, 5.5125] kHz in the text and [2, 5.5125) kHz in the band-index list; the intended inclusion of the upper endpoint should be clarified.
[III-A1, Eq. (14)] Equation (14) for the spectral Renyi entropy is missing parentheses: as written it appears as log(ΣPSD)^q rather than the intended log(ΣPSD^q)/(1−q).
[Whole manuscript] The manuscript contains several typographical errors, including 'contaning', 'conciousness', and '?rst' in Section IV-B, and would benefit from a careful proofreading pass.
[II, IV] No code or data availability statement is provided, which limits the reproducibility of the reported results; a public release of the feature extraction and evaluation scripts would strengthen the paper.

Circularity Check

1 steps flagged · score 6.0 of 10

Feature selection in §III-B/§IV-A is fit to 50% of all observations before the 5-fold test partitions are applied, so the reported SEN/SPE/AUC and the claimed superiority over [15], [17], [18] are not fully out-of-sample predictions.

fitted input called prediction [Section III-B (feature selection) and Section IV-A, applied before the cross-validation partitions in Section III-D; results in Tables IV/V and the abstract.]
"10% of the observations of the feature space were randomly selected for each part of the protocol. ... The Relieff algorithm [32] - a widely used supervised feature selection algorithm for two-class problems - was applied to the selected observations in order to identify the best 29 features in each part. ... To ensure the generalisation capabilities of the proposed feature selection approach, we carried out the process described in Section III-B for five times. The five randomly selected groups were disjoint sets, so 50% of observations were employed in this step. ..."

The 29-feature set is selected by a supervised algorithm (Relieff) on 10% of observations per protocol part, repeated on five disjoint random subsets totaling 50% of all observations, before the block-wise 5-fold train/test partitions of §III-D are formed. Those folds evaluate on the full dataset, so the random selection subsets overlap with observations that later appear in test folds. The paper's only safeguard states that no long-term features from the train or test sets were used for selection, but the selection is explicitly at short-term level, so this does not prevent short-term test-frame labels from influencing the chosen features. Thus the reported SEN, SPE, and AUC in Tables IV/V and the abstract are not independent predictions: the feature set is itself fitted to test labels.

full rationale

The paper's central claim is that the proposed AvgSD/SVM pipeline outperforms prior cough detectors, supported by Table IV/V and the abstract's 92.71% sensitivity, 88.58% specificity, and 90.69% AUC. The pipeline's only fitted component before classification is the supervised feature-selection step: Relieff selects 29 short-term features from 50% of all observations (§III-B, §IV-A). The 5-fold cross-validation partitions described in §III-D are then applied to the whole dataset, with no statement that the selection was nested inside training folds. The text's disclaimer about 'long-term features from the train or test sets' does not address this, because the selection operates on short-term features. Therefore, the selected feature set is informed by test labels, and the reported performance is not a clean out-of-sample evaluation. This is a concrete fitted-input-called-prediction circularity affecting the main superiority claim. No other load-bearing circular steps were found: the use of [18] as a baseline is a reimplementation on the same database rather than an argument from self-citation, and no uniqueness theorem or ansatz is smuggled in via citation.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim is empirical, so the ledger lists hand-chosen hyperparameters and domain assumptions rather than mathematical axioms. The most important items are the annotation-quality assumption and the noise-transfer assumption used in feature selection. No new physical entities are introduced.

free parameters (6)

Number of frequency bands = 5
Chosen after inspecting the average periodogram of cough events, with band edges at 0-0.5, 0.5-1, 1-1.5, 1.5-2, 2-5.5125 kHz. This is a hand-chosen, data-informed design.
Short-term frame length and overlap = 75 ms frame, 19 ms overlap
Justified by spectral estimation needs and cough phase duration; the text says '19s overlap', likely a typo for 19 ms. A modeling choice.
Renyi entropy order q = 4
Set for this work without further justification in Section III-A1.
Number of selected features = 29
Based on MLE intrinsic dimension estimates of 25-30 from the data, Section III-B.
SVM kernel degree = 2 (polynomial)
Chosen without a reported search, Section III-D.
BoAW codebook size per class (Kpos/Kneg) = 16
Some values (16/32, 32/16, 32/32) were tested, and AvgSD still performed better; not a free parameter for the main system.

assumptions (4)

domain assumption Audio signals are linear and stationary within short frames of 20-100 ms.
Stated in Section I as the basis of short-term processing.
domain assumption Manual frame annotations are accurate ground truth.
Section II describes manual annotation with no inter-annotator agreement or validation.
ad hoc to paper If a feature is a good descriptor in noisy environments, it will also be good in more favorable conditions.
Section III-B, the combination procedure assumes noise-robust features transfer to clean conditions.
domain assumption The predefined five frequency bands capture the cough spectral structure.
Section III-A1, based on the average periodogram of cough events.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Machine Hearing System for Robust Cough Detection Based on a High-Level Representation of Band-Specific Audio Features." pith.science (2026). https://pith.science/paper/DHAWEBY5

@misc{pith2026241201996,
  author       = {Pith},
  title        = {Pith review of: A Machine Hearing System for Robust Cough Detection Based on a High-Level Representation of Band-Specific Audio Features},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DHAWEBY5}},
  note         = {Machine review of arXiv:2412.01996}
}

read the original abstract

Cough is a protective reflex conveying information on the state of the respiratory system. Cough assessment has been limited so far to subjective measurement tools or uncomfortable (i.e., non-wearable) cough monitors. This limits the potential of real-time cough monitoring to improve respiratory care. Objective: This paper presents a machine hearing system for audio-based robust cough segmentation that can be easily deployed in mobile scenarios. Methods: Cough detection is performed in two steps. First, a short-term spectral feature set is separately computed in five predefined frequency bands: [0, 0.5), [0.5, 1), [1, 1.5), [1.5, 2), and [2, 5.5125] kHz. Feature selection and combination are then applied to make the short-term feature set robust enough in different noisy scenarios. Second, high-level data representation is achieved by computing the mean and standard deviation of short-term descriptors in 300 ms long-term frames. Finally, cough detection is carried out using a support vector machine trained with data from different noisy scenarios. The system is evaluated using a patient signal database which emulates three real-life scenarios in terms of noise content. Results: The system achieves 92.71% sensitivity, 88.58% specificity, and 90.69% Area Under Receiver Operating Characteristic (ROC) curve (AUC), outperforming state-of-the-art methods. Conclusion: Our research outcome paves the way to create a device for cough monitoring in real-life situations. Significance: Our proposal is aligned with a more comfortable and less disruptive patient monitoring, with benefits for patients (allows self-monitoring of cough symptoms), practitioners (e.g., assessment of treatments or better clinical understanding of cough patterns), and national health systems (by reducing hospitalizations).

Figures

Figures reproduced from arXiv: 2412.01996 by the authors.

**Figure 1.** Representation of different cough events and their spectrograms where the specific phases have been detailed: (I) explosive phase, (II) intermediate [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗

**Figure 2.** Patient-aggregated SNR distribution for each part of the protocol [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Each block is described in the following subsections. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (9 more)

**Figure 3.** Figure 3: Processing pipeline of the proposed cough detection system with specific references to the sections describing each block. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png]

**Figure 4.** Figure 4: Sample average periodogram of the recorded cough events. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: Pipeline of the the feature selection process. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Explanatory diagram of how high-level representation is obtained from short-term features. In the AvgSD method, feature-wise mean and standard [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: Representation of the five block-division train-test partitions employed [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

**Figure 8.** Figure 8: Protocol-averaged mean ROC curves for all the compared methods. [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]

**Figure 9.** Figure 9: Classification results (%) obtained from leave-one patient-out cross-validation of each part of the acquisition protocol. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]

**Figure 10.** Figure 10: Classification results (%) obtained from leave-one patient-out cross-validation for the final system: (a) when a model is built for each part of the [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

**Figure 11.** Figure 11: Illustration of detected (right) and missed (left) cough events. [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 42 canonical work pages

[15]

Cough detection by ensembling multiple frequency subband features,

M. You et al. , “Cough detection by ensembling multiple frequency subband features,” Biomedical Signal Processing and Control , vol. 33, pp. 132 – 140, 2017

work page 2017
[17]

Deep neural networks for identifying cough sounds,

J. Amoh and K. Odame, “Deep neural networks for identifying cough sounds,” IEEE Transactions on Biomedical Circuits and Systems, vol. 10, no. 5, pp. 1003–1011, Oct 2016

work page 2016
[18]

Robust detection of audio-cough events using local hu moments,

J. Monge-Alvarez et al., “Robust detection of audio-cough events using local hu moments,” IEEE Journal of Biomedical and Health Informatics, vol. PP, no. 99, pp. 1–1, 2018

work page 2018
[1]

What is cough and what should be measured?

G. A. Fontana and J. Widdicombe, “What is cough and what should be measured?” Pulmonary Pharmacology & Therapeutics , vol. 20, no. 4, pp. 307 – 312, 2007, special Issue: Fourth International Symposium on Cough

work page 2007
[2]

Cough in exercise and athletes,

J. Hull, J. Dickinson, and A. Jackson, “Cough in exercise and athletes,” Pulmonary Pharmacology & Therapeutics , vol. 47, pp. 49 – 55, 2017, cough hypersensitivity syndrome: mechanisms and clinical approaches

work page 2017
[3]

Semantics and types of cough,

K. F. Chung et al. , “Semantics and types of cough,” Pulmonary Pharmacology & Therapeutics , vol. 22, no. 2, pp. 139 – 142, 2009, the 2008 Fifth International Symposium on Cough

work page 2009
[4]

Ers guidelines on the assessment of cough,

A. H. Morice et al. , “Ers guidelines on the assessment of cough,” European Respiratory Journal, vol. 29, no. 6, pp. 1256–1276, 2007

work page 2007
[5]

How best to measure cough clinically,

S. S. Birring and A. Spinou, “How best to measure cough clinically,” Current Opinion in Pharmacology , vol. 22, pp. 37 – 40, 2015

work page 2015

Show all 43 references

[6]

Evaluation of a cough-specific quality-of-life questionnaire,

C. T. French et al. , “Evaluation of a cough-specific quality-of-life questionnaire,” Chest, vol. 121, no. 4, pp. 1123 – 1131, 2002

2002
[7]

Measurement of cough,

K. Chung, “Measurement of cough,” Respiratory Physiology & Neuro- biology, vol. 152, no. 3, pp. 329 – 339, 2006, cough and its Regulation

2006
[8]

Quality of life and psychosocial aspects of cough,

K. Brignall, B. Jayaraman, and S. S. Birring, “Quality of life and psychosocial aspects of cough,” Lung, vol. 186, no. 1, pp. 55–58, Feb 2008

2008
[9]

New developments in the objective assess- ment of cough,

J. Smith and A. Woodcock, “New developments in the objective assess- ment of cough,” Lung, vol. 186, no. 1, pp. 48–54, Feb 2008

2008
[10]

Automatic cough segmentation from non-contact sound recordings in pediatric wards,

Y . A. Amrulloh et al., “Automatic cough segmentation from non-contact sound recordings in pediatric wards,” Biomedical Signal Processing and Control, vol. 21, pp. 126 – 136, 2015

2015
[11]

Evaluation of an ambulatory system for the quantification of cough frequency in patients with chronic obstructive pulmonary disease,

M. A. Coyle et al. , “Evaluation of an ambulatory system for the quantification of cough frequency in patients with chronic obstructive pulmonary disease,” Cough, vol. 1, no. 1, p. 3, Aug 2005

2005
[12]

Ambulatory methods for recording cough,

J. Smith, “Ambulatory methods for recording cough,” Pulmonary Phar- macology & Therapeutics , vol. 20, no. 4, pp. 313 – 318, 2007, special Issue: Fourth International Symposium on Cough

2007
[13]

B. W. Schuller, Intelligent Audio Analysis. Springer Berlin Heidelberg, 2013

2013
[14]

Detection of cough signals in continuous audio record- ings using hidden markov models,

S. Matos et al., “Detection of cough signals in continuous audio record- ings using hidden markov models,” IEEE Transactions on Biomedical Engineering, vol. 53, no. 6, pp. 1078–1083, June 2006

2006
[16]

Audio-based event detection in office live environments using optimized mfcc-svm approach,

S. E. Kkbay and M. Sert, “Audio-based event detection in office live environments using optimized mfcc-svm approach,” in Proceedings of the 2015 IEEE 9th International Conference on Semantic Computing (IEEE ICSC 2015) , Feb 2015, pp. 475–480

2015
[19]

Reliable detection of audio events in highly noisy environments,

P. Foggia et al. , “Reliable detection of audio events in highly noisy environments,” Pattern Recognition Letters, vol. 65, pp. 22 – 28, 2015

2015
[20]

Bag-of-features methods for acoustic event detection and classification,

R. Grzeszick, A. Plinge, and G. A. Fink, “Bag-of-features methods for acoustic event detection and classification,” IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 25, no. 6, pp. 1242–1252, June 2017

2017
[21]

Gaussian mixture modeling of short- time fourier transform features for audio fingerprinting,

A. Ramalingam and S. Krishnan, “Gaussian mixture modeling of short- time fourier transform features for audio fingerprinting,” IEEE Transac- tions on Information Forensics and Security , vol. 1, no. 4, pp. 457–463, Dec 2006

2006
[22]

Regional analysis of spontaneous meg rhythms in patients with alzheimer’s disease using spectral entropies,

J. Poza et al., “Regional analysis of spontaneous meg rhythms in patients with alzheimer’s disease using spectral entropies,” Annals of Biomedical Engineering, vol. 36, no. 1, pp. 141–152, Jan 2008

2008
[23]

Haykin, Communication Systems, 5th ed

S. Haykin, Communication Systems, 5th ed. Wiley Publishing, 2009

2009
[24]

Chapter 4 - audio features,

T. Giannakopoulos and A. Pikrakis, “Chapter 4 - audio features,” in Introduction to Audio Analysis, T. Giannakopoulos and A. Pikrakis, Eds. Oxford: Academic Press, 2014, pp. 59 – 103

2014
[25]

Application of tonal index to pulmonary wheezes detection in asthma monitoring,

M. Wisniewski and T. P. Zielinski, “Application of tonal index to pulmonary wheezes detection in asthma monitoring,” in 2011 19th European Signal Processing Conference , Aug 2011, pp. 1544–1548

2011
[26]

An overview of applications and advance- ments in automatic sound recognition,

R. V . Sharan and T. J. Moir, “An overview of applications and advance- ments in automatic sound recognition,” Neurocomputing, vol. 200, pp. 22 – 34, 2016

2016
[27]

Low-level descriptors,

H.-G. Kim, N. Moreau, and T. Sikora, “Low-level descriptors,” in MPEG-7 Audio and Beyond. John Wiley & Sons, Ltd, 2006, pp. 13–57

2006
[28]

Joint application of audio spectral envelope and tonality index in an e-asthma monitoring system,

M. Wisniewski and T. P. Zielinski, “Joint application of audio spectral envelope and tonality index in an e-asthma monitoring system,” IEEE Journal of Biomedical and Health Informatics , vol. 19, no. 3, pp. 1009– 1018, May 2015

2015
[29]

Robust speech recognition in noisy environments based on subband spectral centroid histograms,

B. Gajic and K. K. Paliwal, “Robust speech recognition in noisy environments based on subband spectral centroid histograms,” IEEE Transactions on Audio, Speech, and Language Processing, vol. 14, no. 2, pp. 600–608, March 2006

2006
[30]

Feature extraction and selection,

G. Dougherty, “Feature extraction and selection,” in Pattern Recognition and Classification: An Introduction . New York, NY: Springer New York, 2013, pp. 123–141

2013
[31]

Maximum likelihood estimation of intrinsic dimension,

E. Levina and P. J. Bickel, “Maximum likelihood estimation of intrinsic dimension,” in Advances in Neural Information Processing Systems 17 , L. K. Saul, Y . Weiss, and L. Bottou, Eds. MIT Press, 2005, pp. 777– 784

2005
[32]

Theoretical and empirical anal- ysis of relieff and rrelieff,

M. Robnik-Sikonja and I. Kononenko, “Theoretical and empirical anal- ysis of relieff and rrelieff,” Machine Learning, vol. 53, no. 1, pp. 23–69, Oct 2003

2003
[33]

Riley, E

M. Riley, E. Heinen, and J. Ghosh, A text retrieval approach to content- based audio retrieval , 2008, pp. 295–300

2008
[34]

Bag-of-audio-words approach for multi- media event classification,

S. Pancoast and M. Akbacak, “Bag-of-audio-words approach for multi- media event classification,” 2012

2012
[35]

A bag-of-features approach to acoustic event detection,

A. Plinge, R. Grzeszick, and G. A. Fink, “A bag-of-features approach to acoustic event detection,” in 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , May 2014, pp. 3704–3708

2014
[36]

Algorithm as 136: A k-means clustering algorithm,

J. A. Hartigan and M. A. Wong, “Algorithm as 136: A k-means clustering algorithm,” Journal of the Royal Statistical Society. Series C (Applied Statistics) , vol. 28, no. 1, pp. 100–108, 1979

1979
[37]

Giannakopoulos and A

T. Giannakopoulos and A. Pikrakis, Introduction to audio analysis: a MATLAB(R) approach. Academic Press, 2014

2014
[38]

Note on the sampling error of the difference between correlated proportions or percentages,

Q. McNemar, “Note on the sampling error of the difference between correlated proportions or percentages,” Psychometrika, vol. 12, no. 2, pp. 153–157, 1947

1947
[39]

I-vector modeling of speech attributes for automatic foreign accent recognition,

H. Behravan et al., “I-vector modeling of speech attributes for automatic foreign accent recognition,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , vol. 24, no. 1, pp. 29–41, 2015

2015
[40]

Selfmanagement education and regular practitioner review for adults with asthma,

P. G. Gibson, “Selfmanagement education and regular practitioner review for adults with asthma,” Jul 2002

2002
[41]

The smartphone as a medical device: Assessing enablers, benefits and challenges,

E. Agu et al., “The smartphone as a medical device: Assessing enablers, benefits and challenges,” in 2013 IEEE International Conference on Sensing, Communications and Networking (SECON) , June 2013, pp. 76–80

2013
[42]

Cough and environmental air pollution in china,

Q. Zhang et al. , “Cough and environmental air pollution in china,” Pulmonary Pharmacology & Therapeutics, vol. 35, pp. 132 – 136, 2015

2015
[43]

Cough management in primary, secondary and tertiary settings,

K. Wang et al., “Cough management in primary, secondary and tertiary settings,” Pulmonary Pharmacology & Therapeutics , vol. 47, pp. 93 – 98, 2017, cough hypersensitivity syndrome: mechanisms and clinical approaches

2017

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [15]

Cough detection by ensembling multiple frequency subband features,

M. You et al. , “Cough detection by ensembling multiple frequency subband features,” Biomedical Signal Processing and Control , vol. 33, pp. 132 – 140, 2017

work page 2017

[2] [17]

Deep neural networks for identifying cough sounds,

J. Amoh and K. Odame, “Deep neural networks for identifying cough sounds,” IEEE Transactions on Biomedical Circuits and Systems, vol. 10, no. 5, pp. 1003–1011, Oct 2016

work page 2016

[3] [18]

Robust detection of audio-cough events using local hu moments,

J. Monge-Alvarez et al., “Robust detection of audio-cough events using local hu moments,” IEEE Journal of Biomedical and Health Informatics, vol. PP, no. 99, pp. 1–1, 2018

work page 2018

[4] [1]

What is cough and what should be measured?

G. A. Fontana and J. Widdicombe, “What is cough and what should be measured?” Pulmonary Pharmacology & Therapeutics , vol. 20, no. 4, pp. 307 – 312, 2007, special Issue: Fourth International Symposium on Cough

work page 2007

[5] [2]

Cough in exercise and athletes,

J. Hull, J. Dickinson, and A. Jackson, “Cough in exercise and athletes,” Pulmonary Pharmacology & Therapeutics , vol. 47, pp. 49 – 55, 2017, cough hypersensitivity syndrome: mechanisms and clinical approaches

work page 2017

[6] [3]

Semantics and types of cough,

K. F. Chung et al. , “Semantics and types of cough,” Pulmonary Pharmacology & Therapeutics , vol. 22, no. 2, pp. 139 – 142, 2009, the 2008 Fifth International Symposium on Cough

work page 2009

[7] [4]

Ers guidelines on the assessment of cough,

A. H. Morice et al. , “Ers guidelines on the assessment of cough,” European Respiratory Journal, vol. 29, no. 6, pp. 1256–1276, 2007

work page 2007

[8] [5]

How best to measure cough clinically,

S. S. Birring and A. Spinou, “How best to measure cough clinically,” Current Opinion in Pharmacology , vol. 22, pp. 37 – 40, 2015

work page 2015