REVIEW 4 major objections 4 minor 25 references
Deciphering GunType Hierarchy through Acoustic Analysis of Gunshot Recordings
T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read A joint gunshot-detection and firearm-classification CNN achieves a mean average precision of 0.58 on clean curated audio, far above an SVM baseline of 0.39, while degrading sharply on noisy web audio.
desk verdict A plausible empirical study whose headline mAP is uninterpretable until the 'clean' corpus is shown to contain real gunshots rather than sound-effects-library recordings. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is the joint detection-and-classification CNN operating on log mel spectrograms. Shared convolutional layers learn low-level time-frequency patterns common to gunshot events; the detection branch uses a sigmoid output for gunshot presence, and the classification branch uses a softmax output over the five firearm categories, applied when the detection branch positively identifies a gunshot. The underlying physical basis is the two-component gunshot acoustic structure—the roughly 3 to 5 millisecond muzzle blast and, for supersonic rounds, the 200 to 400 microsecond ballistic shockwave—whose timing, spectral content, and presence vary by firearm type, ammunition, and shooting direction.
What would settle it
Take a set of firearms of known make and model, record live shots with several phones at different distances, and run the published CNN on those clips without retraining; if per-class mean average precision falls to near chance, or far below 0.58, while the SVM baseline holds up, the paper's central claim about learnable acoustic gun-type hierarchies is disproved.
Extended reading notes
Core claim
On its own terms, the central discovery is that a CNN with shared convolutional layers and two task-specific output branches—one binary gunshot/no-gunshot detector and one five-way firearm classifier—extracts enough structure from 128-band log mel spectrograms to classify gun type with a mean average precision of 0.58 on a curated set of 3459 recordings, compared with 0.39 for an SVM baseline. The gunshot-detection branch reaches an F1 score of 0.86 for the gunshot class, and the most reliable firearm categories are submachine guns and machine guns, whose rapid-fire temporal signatures are acoustically distinctive, while rifles and shotguns remain harder to separate.
Load-bearing premise
The results stand or fall on whether the 3459 curated clips—drawn from internal collections, public playlists, and commercial sound-effect libraries—are correctly labeled by firearm type and acoustically faithful to real gunshots as heard on phones and field microphones.
Editorial extensions
If this is right
- If the 0.58 mean average precision result reproduces, a commodity-device gunshot detection system could replace expensive fixed infrastructure, such as the roughly $65,000 per square mile per year cited in the paper, with distributed cell-phone microphones.
- Gun type labels from audio could give first responders tactical information—distinguishing a handgun from a rifle or automatic weapon—before visual confirmation.
- The sharp drop to 0.35 mean average precision on web-sourced audio implies that noise robustness and domain adaptation, not raw architecture, are the next bottleneck for real-world use.
- The joint architecture's gunshot-detection branch, with F1 0.86 for gunshot events, can serve as a gate to reduce downstream classification errors in a deployed alerting system.
Reading between the lines
- The paper leaves implicit that its five-class hierarchy is broad; a system aimed at forensic identification or court admissibility would need per-caliber or per-model discrimination, which the current data and labels do not support.
- The acoustic hierarchy it reports—muzzle blast duration and peak frequency separating shotguns, rifles, and handguns—could be tested on large uncontrolled sound-event datasets to see whether the same features transfer outside curated gunshot libraries.
- A testable extension would measure mean average precision as a function of distance-to-microphone and recording device, since the clean curated data does not report those variables.
- The reliance on sound-effect libraries as training data suggests that matching the recording conditions of the target deployment, rather than adding more gunshot clips, may be the fastest way to close the 0.58-to-0.35 gap.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper presents a study on gunshot detection and firearm type classification from audio recordings. The authors curate a dataset of 3459 recordings across five classes (Rifle, Submachine Gun, Handgun/Pistol, Machine Gun, Shotgun) from sources including a CMU internal collection, Spotify playlists, and a commercial sound-effects library. They extract log mel spectrograms, train an SVM baseline and a multi-task CNN for joint detection and classification, and report results on clean and noisy web data. The central claim is that the CNN achieves a mean average precision (mAP) of 0.58 on clean labeled data, substantially outperforming the SVM baseline (mAP 0.39), while performance drops to mAP 0.35 on noisy web-sourced data.
Significance. If the results are reliable, the paper is a useful empirical contribution to low-cost acoustic gunshot classification: it demonstrates a concrete deep-learning pipeline on a relatively large curated corpus, identifies which firearm categories are acoustically separable (Submachine Gun, Machine Gun) and which are confusable (Shotgun, Rifle), and quantifies a substantial domain shift to noisy web data. The paper does not ship code or data, however, and several load-bearing experimental details are unspecified, so the quantitative claims are currently hard to verify. The core idea—combining detection and classification in a joint model—is sensible and the comparison to an SVM baseline is appropriate in spirit, even though the baseline is under-specified.
major comments (4)
- [§4.2.1 and §4.4] The SVM baseline is not adequately specified. The text states that features were 'likely derived' from the audio representations and that the SVM was 'configured for multi-class classification,' but no exact feature set, kernel, hyperparameters, or training procedure is given. Without this information, the reported mAP of 0.39 for the SVM cannot be interpreted or reproduced, and the claim that the CNN 'significantly outperforms' the baseline is not fully supported. Please provide a precise description of the SVM features and configuration, or explicitly reframe the SVM result as an approximate baseline.
- [§5.2, Table 4] The 'Relevant' metric is defined only by an assumption: the text says 'we will assume' that Relevant scores are conditioned on correct gunshot detection. This is a load-bearing ambiguity because the per-class F1 scores in Table 4 are discussed mainly through the 'Relevant' column. Please define the exact computation (e.g., whether it uses the detection branch's hard decision, a confidence threshold, or ground-truth detection masks) and report the number of test instances in each subset. Also clarify how 'Overall' scores differ from 'Relevant' scores.
- [§3 and §4.3] The provenance and validity of the curated dataset are not established. Recordings are drawn from Spotify playlists and a commercial Firearm Sound Effects Library, which are largely sound effects rather than verified live gunshot recordings, and the labels come from manual review. No per-file provenance, label-noise estimate, or inter-annotator agreement is provided. In addition, the 60/20/20 random split in §4.3 does not separate by source, so clips from the same playlist or library can appear in both training and test sets; the model may then learn source-specific artifacts (e.g., reverb, mastering, synthetic signatures) rather than firearm acoustics. Please provide a source-disjoint split, per-source performance analysis, and some validation of label correctness.
- [§5.3] The reported mAP values (0.58 for CNN, 0.39 for SVM, 0.35 for noisy web data) are given without error bars, confidence intervals, or significance tests. The paper mentions 5-fold cross-validation in §4.3, but no variance across folds is reported. The noisy web data evaluation is also under-described: the number of clips, their sources, and how labels were obtained are not stated, making the 0.35 mAP difficult to interpret. Please report fold-level statistics and a detailed description of the web evaluation set.
minor comments (4)
- [§5.1, Table 3] The sentence in the text ends with 'it is correct 82' and omits the percentage sign and the rest of the sentence; the intended value is presumably 82%.
- [§2, Table 1] Table 1 lists acoustic characteristics (peak frequency, duration, SPL, shockwave) but no citation or measurement methodology is given for these numbers. Please indicate whether the values are from the cited literature or from the authors' own measurements, and, if the latter, how they were obtained.
- [§4.1] The description of the BoAW feature is vague ('a dictionary of acoustic words ... was considered'). Since the BoAW features are not used in the reported experiments, please state explicitly that they were not used in the final evaluation, or provide details if they were.
- [§5.2] The discussion of per-class performance describes Handgun/Pistol as achieving a 'moderate F1 score (0.54 relevant)' while Table 4 reports 0.64 in the Relevant column; please ensure consistency between the text and the table.
Circularity Check
No significant circularity: the reported mAPs are empirical results from a held-out evaluation of models trained on a self-curated dataset, not derivations from fitted parameters or from self-citations.
full rationale
The central claims, CNN mAP 0.58 vs SVM mAP 0.39 on clean data, are measured performances under a standard 60/20/20 split with 5-fold cross-validation (Section 4.3). The input features (log mel spectrograms, Section 4.1) and the class labels (manual annotation, Section 3) are external to the model; no fitted parameter is renamed as a prediction, and no equation equates the output to the input by construction. Table 1 is presented as acoustic background informed by prior literature, not as fitted inputs to the classifier. The self-citations ([15], [20]-[22]) support methodological choices such as the joint architecture and weak-label/web-data approaches, but they do not serve as the load-bearing evidence for the reported evaluation; removing them would not change the measured mAPs or the SVM-versus-CNN comparison. Concerns about dataset provenance, possible sound-effect-library artifacts, and lack of external validation are real validity and generalization risks, but they are not circularity: the paper does not claim to predict anything that was itself used as a fitting target. The admission of label noise in Section 7 is a limitation statement, not a circular step. Overall, no significant circularity is present; the minor self-citations are not load-bearing, so the appropriate score is 1.
Assumptions & free parameters
assumptions (3)
- domain assumption The curated recording labels are accurate.
- domain assumption Recordings from Spotify playlists and commercial sound-effect libraries are representative of real gunshots.
- domain assumption Acoustic characteristics in Table 1 differentiate firearm types as described.
Cite this review
Pith. "Pith review of Deciphering GunType Hierarchy through Acoustic Analysis of Gunshot Recordings." pith.science (2026). https://pith.science/paper/EIXVSZ22
@misc{pith2026250620609,
author = {Pith},
title = {Pith review of: Deciphering GunType Hierarchy through Acoustic Analysis of Gunshot Recordings},
year = {2026},
howpublished = {\url{https://pith.science/paper/EIXVSZ22}},
note = {Machine review of arXiv:2506.20609}
}
read the original abstract
The escalating rates of gun-related violence and mass shootings represent a significant threat to public safety. Timely and accurate information for law enforcement agencies is crucial in mitigating these incidents. Current commercial gunshot detection systems, while effective, often come with prohibitive costs. This research explores a cost-effective alternative by leveraging acoustic analysis of gunshot recordings, potentially obtainable from ubiquitous devices like cell phones, to not only detect gunshots but also classify the type of firearm used. This paper details a study on deciphering gun type hierarchies using a curated dataset of 3459 recordings. We investigate the fundamental acoustic characteristics of gunshots, including muzzle blasts and shockwaves, which vary based on firearm type, ammunition, and shooting direction. We propose and evaluate machine learning frameworks, including Support Vector Machines (SVMs) as a baseline and a more advanced Convolutional Neural Network (CNN) architecture for joint gunshot detection and gun type classification. Results indicate that our deep learning approach achieves a mean average precision (mAP) of 0.58 on clean labeled data, outperforming the SVM baseline (mAP 0.39). Challenges related to data quality, environmental noise, and the generalization capabilities when using noisy web-sourced data (mAP 0.35) are also discussed. The long-term vision is to develop a highly accurate, real-time system deployable on common recording devices, significantly reducing detection costs and providing critical intelligence to first responders.
Reference graph
Works this paper leans on
-
[1]
Gun Violence Archive, “Mass shootings database,” 2024. [Online]. Available: https://www.gunviolencearchive.org
work page 2024
-
[2]
ShotSpotter ® Gunshot Detection System,
SoundThinking, Inc., “ShotSpotter ® Gunshot Detection System,” https: //www.shotspotter.com/, 2025, accessed: 2025-05-02. Company rebranded to SoundThinking in 2023
work page 2025
-
[3]
Latest achievements in gunfire detection systems,
J. Millet and B. Baligand, “Latest achievements in gunfire detection systems,” Proceedings of the of the RTO-MP-SET-107 Battlefield Acoustic Sensing for ISR Applications , 2006
work page 2006
-
[4]
Classification and modeling of acoustic gunshot signatures,
M. Djeddou and T. Touhami, “Classification and modeling of acoustic gunshot signatures,” Arabian journal for science and engineering , vol. 38, no. 12, pp. 3399–3406, 2013
work page 2013
-
[5]
Gun type recognition from gunshot audio recordings,
E. Kiktova, M. Lojka, M. Pleva, J. Juhar, and A. Cizmar, “Gun type recognition from gunshot audio recordings,” in3rd international workshop on biometrics and forensics (IWBF 2015) . IEEE, 2015, pp. 1–6
work page 2015
-
[6]
Sound of guns: digital forensics of gun audio samples meets artificial intelligence,
S. Raponi, G. Oligeri, and I. M. Ali, “Sound of guns: digital forensics of gun audio samples meets artificial intelligence,” Multimedia tools and applications, vol. 81, no. 21, pp. 30 387–30 412, 2022
work page 2022
-
[7]
Measurements, analysis, classification, and detection of gunshot and gunshot-like sounds,
R. B. Singh and H. Zhuang, “Measurements, analysis, classification, and detection of gunshot and gunshot-like sounds,” Sensors, vol. 22, no. 23, p. 9170, 2022
work page 2022
-
[8]
Acoustical characterization of gunshots,
R. C. Maher, “Acoustical characterization of gunshots,” in 2007 IEEE Workshop on Signal Processing Applications for Public Security and Forensics, 2007, pp. 1–5
work page 2007
Show all 25 references
-
[9]
Deciphering gunshot recordings,
R. C. Maher and S. R. Shaw, “Deciphering gunshot recordings,” in Audio Engineering Society Conference: 33rd International Conference: Audio Forensics-Theory and Practice. Audio Engineering Society, 2008
2008
-
[10]
Impulse acoustic event detection, classification, and localization system,
J. Svatos and J. Holub, “Impulse acoustic event detection, classification, and localization system,” IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1–15, 2023
2023
-
[11]
Technical report of the video event reconstruction and analysis (vera) system–shooter localization, models, interface, and beyond,
J. Liang, J. D. Aronson, and A. Hauptmann, “Technical report of the video event reconstruction and analysis (vera) system–shooter localization, models, interface, and beyond,” arXiv preprint arXiv:1905.13313 , 2019
1905 arXiv
-
[12]
Modeling and signal processing of acoustic gunshot recordings,
R. C. Maher, “Modeling and signal processing of acoustic gunshot recordings,” in 2006 IEEE 12th Digital Signal Processing Workshop & 4th IEEE Signal Processing Education Workshop . IEEE, 2006, pp. 257–261
2006
-
[13]
openxbow–introducing the passau open- source crossmodal bag-of-words toolkit,
M. Schmitt and B. Schuller, “openxbow–introducing the passau open- source crossmodal bag-of-words toolkit,” Journal of Machine Learning Research, vol. 18, no. 96, pp. 1–5, 2017
2017
-
[14]
At the border of acoustics and linguistics: Bag-of-audio-words for the recognition of emotions in speech,
M. Schmitt, F. Ringeval, and B. Schuller, “At the border of acoustics and linguistics: Bag-of-audio-words for the recognition of emotions in speech,” 2016
2016
-
[15]
Computational Audition with Imprecise Labels,
A. P. Shah, “Computational Audition with Imprecise Labels,” 2 2025. [Online]. Available: https://kilthub.cmu.edu/articles/thesis/ Computational Audition with Imprecise Labels/28422542
2025
-
[16]
Audio event detection using weakly labeled data,
A. Kumar and B. Raj, “Audio event detection using weakly labeled data,” in Proceedings of the 24th ACM International Conference on Multimedia , ser. MM ’16. New York, NY , USA: Association for Computing Machinery, 2016, p. 1038–1047. [Online]. Available: https://doi.org/10.114...
2016
-
[17]
Ast: Audio spectrogram transformer,
Y . Gong, Y .-A. Chung, and J. Glass, “Ast: Audio spectrogram transformer,” arXiv preprint arXiv:2104.01778 , 2021
2021 arXiv
-
[18]
Co-teaching: Robust training of deep neural networks with extremely noisy labels,
B. Han, Q. Yao, X. Yu, G. Niu, M. Xu, W. Hu, I. Tsang, and M. Sugiyama, “Co-teaching: Robust training of deep neural networks with extremely noisy labels,” Advances in neural information processing systems , vol. 31, 2018
2018
-
[19]
Support vector machine,
S. Suthaharan, “Support vector machine,” in Machine learning models and algorithms for big data classification: thinking with examples for effective learning. Springer, 2016, pp. 207–235
2016
-
[20]
Importance of negative sampling in weak label learning,
A. Shah, F. Tang, Z. Ye, R. Singh, and B. Raj, “Importance of negative sampling in weak label learning,” in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024, pp. 7530–7534
2024
-
[21]
Learning sound events from webly labeled data,
A. Kumar, A. Shah, B. Raj, and A. Hauptmann, “Learning sound events from webly labeled data,” arXiv preprint arXiv:1811.09967 , 2018
2018 arXiv
-
[22]
A closer look at weak label learning for audio events,
A. Shah, A. Kumar, A. G. Hauptmann, and B. Raj, “A closer look at weak label learning for audio events,” arXiv preprint arXiv:1804.09288 , 2018
2018 arXiv
-
[23]
mixup: Beyond empirical risk minimization,
H. Zhang, M. Cisse, Y . N. Dauphin, and D. Lopez-Paz, “mixup: Beyond empirical risk minimization,” arXiv preprint arXiv:1710.09412 , 2017
2017 arXiv
-
[24]
Low cost gunshot detection using deep learning on the raspberry pi,
A. Morehead, L. Ogden, G. Magee, R. Hosler, B. White, and G. Mohler, “Low cost gunshot detection using deep learning on the raspberry pi,” in 2019 IEEE International Conference on Big Data (Big Data) . IEEE, 2019, pp. 3038–3044
2019
-
[25]
Shooter localization using social media videos,
J. Liang, J. D. Aronson, and A. Hauptmann, “Shooter localization using social media videos,” in Proceedings of the 27th ACM International Conference on Multimedia , ser. MM ’19. New York, NY , USA: Association for Computing Machinery, 2019, p. 2280–2283. [Online]. Available: h...
2019
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.