Detecting Gas Vapor Leaks Using Uncalibrated Sensors

A. Enis \c{C}etin; Alex Orailoglu; Chengmo Yang; Diaa Badawi; Sule Ozev; Tuba Ayhan

REVIEW 3 major objections 5 minor 31 references

Detecting Gas Vapor Leaks Using Uncalibrated Sensors

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Deep networks fed raw sensor time series detect VOC and ammonia leaks and identify gas types at 96–98% accuracy even with uncalibrated sensors, and a multiplication-free network matches this while tolerating 67.4% weight pruning.

desk verdict The IR experiment is likely compromised by an unstratified window split; the public-benchmark parts are more trustworthy and the AddNet/pruning results are interesting. read the letter →

arxiv 1908.07619 v1 pith:LON552EY submitted 2019-08-20 eess.SP cs.LG

Diaa Badawi , Tuba Ayhan , Sule Ozev , Chengmo Yang , Alex Orailoglu , A. Enis Çetin This is my paper · ORCID

classification eess.SPcs.LG

keywords gasvaporleakdetectionuncalibratedsensorssensordrifttime-seriesneuralnetworksadditiveGANdiscriminatorelectronicnoseinfraredsensing

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 argues that gas vapor leaks can be detected and gas types identified by feeding raw temporal sensor signals directly into deep neural networks, without calibrated sensors or hand-crafted features. Using infrared recordings of VOC leaks and two chemical sensor-array datasets, the authors report 97–98% leak-detection accuracy, roughly 96% gas-type identification accuracy, and drift-batch accuracies that often beat an SVM ensemble baseline. The paper's central algorithmic object is AddNet, a network that replaces multiplication in its layers with a sign-preserving addition operation, matching ordinary CNN accuracy while using almost no multiplications and tolerating pruning of 67.4% of weight magnitudes. The paper concludes that temporal presentation lets networks learn discriminative features by themselves, avoiding fragile hand-crafted drift compensation.

What carries the argument

The load-bearing object is the multiplication-devoid (md) operation $x \oplus y = \mathrm{sgn}(xy)(|x|+|y|)$, used in place of ordinary dot products in convolutional and dense layers; a per-neuron scaling, in practice absorbed by batch normalization, keeps layer outputs in a workable range, and the signum derivative is approximated by a sharp hyperbolic tangent during backpropagation. The paper also relies on presenting sensor signals as short time windows, randomly cropped during training so the networks become invariant to when a leak starts, and on a two-phase GAN scheme whose generator synthesizes extra minority-class signals to help the discriminator classify under class imbalance. The md operation matters because it preserves the sign of ordinary multiplication and induces the $\ell^1$ norm, which the paper uses to justify why AddNet keeps accuracy under heavy weight-magnitude pruning.

What would settle it

Permute the 50 time samples in each infrared test clip and retrain under the same protocol; if leak/no-leak accuracy stays near 97–98%, temporal ordering is not what the network uses, which would contradict the paper's central conclusion.

Watch

Extended reading notes

Core claim

The paper's discovery claim is that time-series sensor measurements are sufficient input for accurate gas-leak detection and gas identification even when the sensors are uncalibrated, because deep networks learn the discriminative features from the temporal structure itself. On an infrared VOC dataset the networks reach 97–98% accuracy, with AddNet (no dropout) reaching 98.2% in the unbalanced setting while the GAN discriminator reaches 98.1%. On a 16-sensor array exposed to CO, ethylene, and methane, both ConvNet and AddNet reach about 96% average accuracy. On a three-year chemical sensor drift dataset, the MLP and GAN variants beat the SVM ensemble on several later batches. The paper further claims that AddNet, which replaces every dot product except in the last layer with a sign-preserving addition operation, matches regular ConvNet accuracy while tolerating pruning of 67.4% of weight magnitudes, making it suitable for energy-limited embedded monitoring.

Load-bearing premise

The load-bearing premise is that the manual selection of infrared regions and their leak/no-leak labels is accurate ground truth, so the high infrared accuracy reflects the temporal sensor signal rather than visual cues or labeling noise.

Editorial extensions

If this is right

Gas leak monitors could run on low-power, embedded platforms, because AddNet confines ordinary dot products to the final layer and performs most inference with additions.
The same temporal-input recipe should transfer to ChemFET and EIS sensors, since the paper explicitly states that its analysis applies to all sensors that produce time-varying signals.
Unbalanced field data, with few leak recordings and many clean-air recordings, can be handled by using a GAN discriminator as the classifier; the paper's unbalanced infrared experiment holds 98% accuracy with only 50 leak training examples.
Weight storage in AddNet can be reduced to roughly a bit per weight in most layers without severe accuracy loss, because pruning 67.4% of magnitudes still yields about 98% accuracy.
For drift-affected chemical sensors, temporal deep networks outperform the SVM ensemble baseline on several later batches, suggesting that drift compensation need not be a separate preprocessing step.

Reading between the lines

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

An untested corollary is that AddNet's tolerance to magnitude pruning may be a direct consequence of the md operator inducing the $\ell^1$ norm, which biases weights toward sparsity; a layer-by-layer pruning study would test this mechanism.
The paper does not separate visual cues from sensor signals in the infrared data, so a natural check is to run the same classifier on temporal signals extracted from blank sky regions or after masking the visible gas cloud.
The paper's reasoning implies that re-releasing raw time series for the long-term drift dataset would likely improve later-batch results, since the current dataset retains only 8 summary statistics per sensor rather than the full temporal signals.
A transferable prediction is that the same temporal deep-network recipe will work on other drift-prone sensor modalities, such as acoustic or vibration leak detectors, which also produce nonstationary signatures with no reliable decision threshold.

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.

Desk Editor's Note

read the letter

Quick take: this is a plausible engineering paper with one load-bearing flaw in its central IR experiment. The public-dataset results are more credible, and the AddNet/pruning material deserves a look.

What's new: the authors apply their md operator (from prior papers) to IR and e-nose time-series gas-leak detection, propose a GAN discriminator as a classifier for imbalanced gas data, and report that AddNet tolerates pruning 67% of weight magnitudes without much accuracy loss. The architectures are specified in enough detail to reproduce, and using the Vergara and Fonollosa public datasets is a plus. The pruning robustness result is genuinely interesting, and the claim that AddNet matches a full-precision ConvNet on these tasks is well supported for the chemical-sensor datasets.

Soft spots, in proportion:

1. The IR experiment (Task 1) has a serious data-split problem. The dataset is 60k windows drawn from a small number of videos, randomly cropped into 32-length signals. The paper says the data is divided into 'disjoint sets' but never says the split is stratified by video or ROI. Overlapping or consecutive windows from the same video are near-duplicates, so if any appear in both train and test, the 97–98% accuracy measures memorization of video-specific traces, not generalization. This is more damaging than the manual-label concern: even correct labels don't fix the independence violation.

2. No error bars or repeated-run statistics anywhere. Tables II and VI report point estimates, and with only 215 instances in the Fonollosa dataset, that's not enough to trust the 96% numbers.

3. The universal-approximation citation is wrong. Reference [27] is Cybenko's sigmoidal theorem, which does not cover AddNet. That's a minor but real scholarly error.

4. The conclusion overreaches: comparing temporal raw signals (Tasks 1 and 2) against the Vergara dataset, which provides only engineered EMA features, doesn't support the claim that 'temporal presentation' is always better.

The paper does not fake anything and the public-benchmark results look reasonable. But the central IR claim cannot be accepted without a video-level split and preferably a data release. I'd send it to peer review because the questions it raises are important, but the referee should make the authors re-analyze Task 1 with a stratified split and add confidence intervals.

Recommendation: engage with it, but be skeptical of the headline accuracy until the split issue is resolved.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes using temporal signals from IR and chemical sensors for gas-leak detection and gas-type identification, comparing three deep-learning approaches: an additive neural network (AddNet), a GAN discriminator used as a classifier (DiscGAN), and a conventional CNN/MLP baseline. It evaluates these methods on one private infrared dataset (Task 1), on the Fonollosa et al. dynamic gas-mixture dataset (Task 2), and on the Vergara et al. drift dataset with a batch-wise train/test protocol (Task 3). The central claims are that temporal sensor signals yield leak detection accuracies of 97-98% and gas-identification accuracies of about 96%, and that AddNet matches or exceeds the baselines while requiring no multiplications except in the final layer and tolerating pruning of 67.4% of weight magnitudes.

Significance. If the reported results are reliable, the temporal-processing approach and the energy-efficient AddNet are practically relevant for embedded and cyber-physical gas-monitoring systems. The paper's strengths include the use of two public benchmark datasets for tasks 2 and 3, detailed architecture tables, a confusion matrix for the main IR experiment, and a systematic pruning analysis. However, the headline Task 1 result is not yet credible because the private IR dataset appears to be split at the window level rather than the recording level, creating a likely train/test leakage that must be resolved before the 97-98% accuracy claim can be evaluated.

major comments (3)

[Section III-A, Tables II-IV] The train/validation/test split is performed at the level of randomly cropped 32-sample windows, not at the level of video recordings or ROIs. Consecutive 50-sample windows extracted from the same ROI are strongly correlated, and the sentence "We divided our data set into three disjoint sets" does not prevent near-duplicate windows from the same recording appearing in both the training and test partitions. The 97-98% accuracies in Table II may therefore reflect memorization of recording-specific intensity traces rather than generalization to unseen leak events. Please re-run the IR experiments with a split that assigns every window from a given video/ROI to the same fold and report the resulting accuracies.
[Tables II, VI, and VIII] Results are reported as point estimates from single training runs. Dropout, random cropping, and the signum-derivative approximation make training stochastic, so differences such as 97.1% vs. 98.2% in Table II may be within run-to-run variation. Please provide mean and standard deviation (or equivalent intervals) over at least five repeated runs, and state the number of runs used for every reported accuracy.
[Section II-B, Eqs. (1)-(7)] The statement that "[t]he proof of AddNet with linear and/or ReLU activation functions satisfying the universal approximation property over the space of Lebesgue integrable functions can be found in [27]" is not supported by the cited reference, which treats classical sigmoidal neural networks with ordinary dot products and says nothing about the md operator defined in Eq. (1). This is a formal claim about AddNet; it needs either a genuine proof or should be removed and replaced by an explicit statement that the universal approximation property of AddNet is not established here.

minor comments (5)

[Section III-A] The manual selection of ROIs and event designations is described only briefly; please state the criteria used, whether the annotator was blind to the sensor time series, and whether any inter-annotator agreement was measured, because label noise directly affects the reported sensitivities and specificities.
[Section III-A] The sentence "We obtained an accuracy of 99.8% for no-gas data and 99.7% for gas-leak data for a regular ConvNet" does not appear in Table II; please clarify which experimental condition these numbers refer to and reconcile all reported accuracy values.
[Section IV] The conclusion lists Task 2 recognition rates as 96.1-96.5%, whereas Table VI reports averages of 96.6% and 96.1%; please correct this inconsistency.
[Section III-B] The phrase "cross validation with holdout method" is ambiguous; if the validation set is drawn randomly four times, please state whether the same test examples can appear in more than one trial and how this affects the reported 140-example validation total.
[Table IV] Please clarify how the weight compression rate is computed when 32-bit intact weights and 1-bit binarized weights are mixed; the textual description is not sufficient to reproduce the stated compression rates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: empirical claims are benchmarked against external data sets and no prediction reduces to a fitted input.

full rationale

The paper's central claims are empirical evaluations of three network types on two public chemical-sensor data sets (Fonollosa et al. [17], Vergara et al. [11]) plus an in-house IR set; none of the reported accuracies is a fitted parameter renamed as a prediction, and no derived quantity is defined in terms of the effect it is supposed to predict. The AddNet md operator is defined by Eq. (1) in the paper, and the property that it 'preserves the sign of regular multiplication operations [24], [25]' follows immediately from that definition, so the self-citations [24], [25] are not load-bearing; removing them would not change the architecture or the comparisons. The universal-approximation sentence cites Cybenko [27], which is an external mathematical theorem, although its applicability to the md operation is not demonstrated; that is a correctness/evidence concern, not circularity. Concerns about manual ROI labeling or random cropping across video-derived windows are dataset-construction and independence threats, not circular reductions. The external benchmarks make the empirical comparisons self-contained; despite minor self-citations and a possibly over-broad proof citation, there is no circular derivation chain to flag.

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

The empirical claims rest primarily on dataset assumptions and hyperparameters, not on a formal derivation. The most fragile items are the manual IR labels, the unproven universal-approximation citation, and the assumption that synthetic minority-class samples improve generalization.

free parameters (4)

tanh sharpness a in signum derivative approximation = 10 (used in Fig. 4)
Chosen by hand in Section II-B; controls how closely tanh approximates the signum derivative and hence affects AddNet gradient updates.
dropout rate = 50% (Task 1), 20% (Task 3)
Set separately for each task with no sensitivity analysis; affects regularization and reported accuracies.
Gaussian augmentation noise standard deviation = 0.1
Used in Section III-C for the drift dataset; chosen by hand to augment data.
random crop length = 32 (Task 1), 40 (Task 2)
Chosen to enlarge training sets and impose translation invariance; changes the input dimension of the networks.

assumptions (4)

ad hoc to paper AddNet with the md operator satisfies the universal approximation property over Lebesgue integrable functions when used with linear or ReLU activations.
Stated in Section II-B with citation [27], but [27] is Cybenko's theorem for sigmoidal networks and does not cover the md operator; no proof is supplied.
domain assumption Manual region-of-interest selection in infrared videos provides correct leak versus no-leak labels.
Section III-A says 'We manually selected regions of interest...'; no masks, inter-rater validation, or code are released.
domain assumption Randomly cropping each 50-sample signal to 32 or 40 samples preserves the class label of every crop.
Used as augmentation in Sections III-A and III-B; if a crop contains mostly pre-leak samples, noisy labels enter training.
domain assumption Generator-produced synthetic samples in DiscGAN lie close enough to real leak signals to improve classifier generalization.
Section II-C and Table VIII rely on this; if synthetic samples are far from real leak data, sensitivity estimates are optimistic.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Detecting Gas Vapor Leaks Using Uncalibrated Sensors." pith.science (2026). https://pith.science/paper/LON552EY

@misc{pith2026190807619,
  author       = {Pith},
  title        = {Pith review of: Detecting Gas Vapor Leaks Using Uncalibrated Sensors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LON552EY}},
  note         = {Machine review of arXiv:1908.07619}
}

read the original abstract

Chemical and infra-red sensors generate distinct responses under similar conditions because of sensor drift, noise or resolution errors. In this work, we use different time-series data sets obtained by infra-red and E-nose sensors in order to detect Volatile Organic Compounds (VOCs) and Ammonia vapor leaks. We process time-series sensor signals using deep neural networks (DNN). Three neural network algorithms are utilized for this purpose. Additive neural networks (termed AddNet) are based on a multiplication-devoid operator and consequently exhibit energy-efficiency compared to regular neural networks. The second algorithm uses generative adversarial neural networks so as to expose the classifying neural network to more realistic data points in order to help the classifier network to deliver improved generalization. Finally, we use conventional convolutional neural networks as a baseline method and compare their performance with the two aforementioned deep neural network algorithms in order to evaluate their effectiveness empirically.

Figures

Figures reproduced from arXiv: 1908.07619 by the authors.

**Figure 2.** Two infrared images of VOC gas leaks. Red rectangles [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 4.** The derivative of tanh(aw) = a sech2 (aw) as a function of parameter w, with a set to 10. where sech(x) = 2 e x+e−x is the hyperbolic secant function, and a is a hyperparameter indicating how sharp the hyperbolic tangent is. The larger the hyperparameter a is, the closer tanh is to the signum function [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: Time-series data generated by four different sensors under exposure to different type of gases (50 time samples for [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 22 canonical work pages

[27]

Approximation by superpositions of a sigmoidal function,

G. Cybenko, “Approximation by superpositions of a sigmoidal function,” Mathematics of control, signals and systems , vol. 2, no. 4, pp. 303–314, 1989

1989
[1]

Indoor air quality and health,

A. P. Jones, “Indoor air quality and health,” Atmospheric environment, vol. 33, no. 28, pp. 4535–4564, 1999

work page 1999
[2]

Concentrations of volatile organic compounds in indoor air–a review,

S. Brown, M. R. Sim, M. J. Abramson, and C. N. Gray, “Concentrations of volatile organic compounds in indoor air–a review,”Indoor air, vol. 4, no. 2, pp. 123–134, 1994

work page 1994
[3]

Benzene, an experimental multipotential carcinogen: results of the long-term bioassays performed at the Bologna Institute of Oncology

C. Maltoni, A. Ciliberti, G. Cotti, B. Conti, and F. Belpoggi, “Benzene, an experimental multipotential carcinogen: results of the long-term bioassays performed at the Bologna Institute of Oncology.” Environ- mental health perspectives , vol. 82, p. 109, 1989

work page 1989
[4]

The indoor volatile organic compound (VOC) characteristics and source identiﬁcation in a new university campus in Tianjin, China,

J. Kang, J. Liu, and J. Pei, “The indoor volatile organic compound (VOC) characteristics and source identiﬁcation in a new university campus in Tianjin, China,” Journal of the Air & Waste Management Association, vol. 67, no. 6, pp. 725–737, 2017

work page 2017
[5]

Detecting gas vapor leaks through uncalibrated sensor based cps,

D. Badawi, S. Ozev, J. B. Christen, C. Yang, A. Orailoglu, and A. E. C ¸ etin, “Detecting gas vapor leaks through uncalibrated sensor based cps,” in ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2019, pp. 8296–8300

work page 2019
[6]

VOC gas leak detection using pyro-electric infrared sensors,

F. Erden, E. B. Soyer, B. U. Toreyin, and A. E. Cetin, “VOC gas leak detection using pyro-electric infrared sensors,” in Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference on . IEEE, 2010, pp. 1682–1685

work page 2010
[7]

Method, device and system for determining the presence of volatile organic compounds (VOC) in video,

A. E. Cetin and B. U. Toreyin, “Method, device and system for determining the presence of volatile organic compounds (VOC) in video,” Apr. 30 2013, uS Patent 8,432,451

work page 2013

Show all 31 references

[8]

Deﬁnitions and typical examples,

W. G ¨opel and K.-D. Schierbaum, “Deﬁnitions and typical examples,” Sensors: Chemical and Biochemical Sensors-Part I , vol. 2, pp. 1–27, 1991

1991
[9]

Frequency analysis of drift in chemical sensors,

F. Davide, C. Di Natale, M. Holmberg, and F. Winquist, “Frequency analysis of drift in chemical sensors,” in Proceedings of the 1st Italian Conference on Sensors and Microsystems, Rome, Italy , 1996, pp. 150– 154

1996
[10]

Drift counterac- tion with multiple self-organising maps for an electronic nose,

M. Zuppa, C. Distante, P. Siciliano, and K. C. Persaud, “Drift counterac- tion with multiple self-organising maps for an electronic nose,” Sensors and Actuators B: Chemical , vol. 98, no. 2-3, pp. 305–317, 2004

2004
[11]

Chemical gas sensor drift compensation using classiﬁer ensembles,

A. Vergara, S. Vembu, T. Ayhan, M. A. Ryan, M. L. Homer, and R. Huerta, “Chemical gas sensor drift compensation using classiﬁer ensembles,” Sensors and Actuators B: Chemical , vol. 166, pp. 320–329, 2012

2012
[12]

Drift correction for gas sensors using multivariate methods,

T. Artursson, T. Ekl ¨ov, I. Lundstr ¨om, P. M ˚artensson, M. Sj ¨ostr¨om, and M. Holmberg, “Drift correction for gas sensors using multivariate methods,” Journal of chemometrics, vol. 14, no. 5-6, pp. 711–723, 2000

2000
[13]

Detection of sensor poisoning using self-diagnostic gas sensors,

D. E. Williams, G. S. Henshaw, and K. F. Pratt, “Detection of sensor poisoning using self-diagnostic gas sensors,” Journal of the Chemical Society, Faraday Transactions, vol. 91, no. 18, pp. 3307–3308, 1995

1995
[14]

National Institute of Standards and Technology. Ammonia. Accessed: 2019-june-16. [Online]. Available: https://webbook.nist.gov/cgi/cbook. cgi?ID=C7664417&Type=IR-SPEC&Index=1

2019
[15]

Accessed: 2019-june-16

FLIR Systems Australia. Accessed: 2019-june-16. [Online]. Available: http://www.ferret.com.au/latest-videos

2019
[16]

Accessed: 2019-june-6

Infrared Cameras Inc. Accessed: 2019-june-6. [Online]. Available: https://infraredcameras.com/thermal-infrared-media/videos/ 9

2019
[17]

Reservoir computing compensates slow response of chemosensor arrays exposed to fast varying gas concentrations in continuous monitoring,

J. Fonollosa, S. Sheik, R. Huerta, and S. Marco, “Reservoir computing compensates slow response of chemosensor arrays exposed to fast varying gas concentrations in continuous monitoring,” Sensors and Actuators B: Chemical , vol. 215, pp. 618–629, 2015

2015
[18]

Imagenet classiﬁcation with deep convolutional neural networks,

A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classiﬁcation with deep convolutional neural networks,” in Advances in neural infor- mation processing systems , 2012, pp. 1097–1105

2012
[19]

Convolutional networks for images, speech, and time series,

Y . LeCun, Y . Bengioet al., “Convolutional networks for images, speech, and time series,” The handbook of brain theory and neural networks , vol. 3361, no. 10, 1995

1995
[20]

A review of unsupervised feature learning and deep learning for time-series modeling,

M. L ¨angkvist, L. Karlsson, and A. Loutﬁ, “A review of unsupervised feature learning and deep learning for time-series modeling,” Pattern Recognition Letters, vol. 42, pp. 11–24, 2014

2014
[21]

Binaryconnect: Training deep neural networks with binary weights during propagations,

M. Courbariaux, Y . Bengio, and J.-P. David, “Binaryconnect: Training deep neural networks with binary weights during propagations,” in Advances in neural information processing systems , 2015, pp. 3123– 3131

2015
[22]

XNOR-Net: ImagenNet classiﬁcation using binary convolutional neural networks,

M. Rastegari, V . Ordonez, J. Redmon, and A. Farhadi, “XNOR-Net: ImagenNet classiﬁcation using binary convolutional neural networks,” in European Conference on Computer Vision. Springer, 2016, pp. 525– 542

2016
[23]

Binarized neural networks: Training deep neural networks with weights and activations constrained to+ 1 or-1,

M. Courbariaux, I. Hubara, D. Soudry, R. El-Yaniv, and Y . Ben- gio, “Binarized neural networks: Training deep neural networks with weights and activations constrained to+ 1 or-1,” arXiv preprint arXiv:1602.02830, 2016

2016 arXiv
[24]

Multiplication free neural network for cancer stem cell detection in h-and-e stained liver images,

D. Badawi, E. Akhan, M. Mallah, A. ¨Uner, R. C ¸ etin-Atalay, and A. E. C ¸ etin, “Multiplication free neural network for cancer stem cell detection in h-and-e stained liver images,” in Compressive Sensing VI: From Diverse Modalities to Big Data Analytics , vol. 10211. Interna...

2017
[25]

Non-euclidean vector product for neural networks,

A. Afrasiyabi, D. Badawi, B. Nasir, O. Yıldız, F. T. Yarman-Vural, and A. E. C ¸ etin, “Non-euclidean vector product for neural networks,” in 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2018, pp. 6862–6866

2018
[26]

Batch normalization: Accelerating deep network training by reducing internal covariate shift,

S. Ioffe and C. Szegedy, “Batch normalization: Accelerating deep network training by reducing internal covariate shift,” arXiv preprint arXiv:1502.03167, 2015

2015 arXiv
[28]

Generative adversarial nets,

I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” in Advances in neural information processing systems , 2014, pp. 2672– 2680

2014
[29]

Unsupervised representation learning with deep convolutional generative adversarial networks,

A. Radford, L. Metz, and S. Chintala, “Unsupervised representation learning with deep convolutional generative adversarial networks,”arXiv preprint arXiv:1511.06434, 2015

2015 arXiv
[30]

Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude,

T. Tieleman and G. Hinton, “Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude,” COURSERA: Neural networks for machine learning , vol. 4, no. 2, pp. 26–31, 2012

2012
[31]

Dropout: a simple way to prevent neural networks from over- ﬁtting,

N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhut- dinov, “Dropout: a simple way to prevent neural networks from over- ﬁtting,” The Journal of Machine Learning Research , vol. 15, no. 1, pp. 1929–1958, 2014

1929

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [27]

Approximation by superpositions of a sigmoidal function,

G. Cybenko, “Approximation by superpositions of a sigmoidal function,” Mathematics of control, signals and systems , vol. 2, no. 4, pp. 303–314, 1989

1989

[2] [1]

Indoor air quality and health,

A. P. Jones, “Indoor air quality and health,” Atmospheric environment, vol. 33, no. 28, pp. 4535–4564, 1999

work page 1999

[3] [2]

Concentrations of volatile organic compounds in indoor air–a review,

S. Brown, M. R. Sim, M. J. Abramson, and C. N. Gray, “Concentrations of volatile organic compounds in indoor air–a review,”Indoor air, vol. 4, no. 2, pp. 123–134, 1994

work page 1994

[4] [3]

Benzene, an experimental multipotential carcinogen: results of the long-term bioassays performed at the Bologna Institute of Oncology

C. Maltoni, A. Ciliberti, G. Cotti, B. Conti, and F. Belpoggi, “Benzene, an experimental multipotential carcinogen: results of the long-term bioassays performed at the Bologna Institute of Oncology.” Environ- mental health perspectives , vol. 82, p. 109, 1989

work page 1989

[5] [4]

The indoor volatile organic compound (VOC) characteristics and source identiﬁcation in a new university campus in Tianjin, China,

J. Kang, J. Liu, and J. Pei, “The indoor volatile organic compound (VOC) characteristics and source identiﬁcation in a new university campus in Tianjin, China,” Journal of the Air & Waste Management Association, vol. 67, no. 6, pp. 725–737, 2017

work page 2017

[6] [5]

Detecting gas vapor leaks through uncalibrated sensor based cps,

D. Badawi, S. Ozev, J. B. Christen, C. Yang, A. Orailoglu, and A. E. C ¸ etin, “Detecting gas vapor leaks through uncalibrated sensor based cps,” in ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2019, pp. 8296–8300

work page 2019

[7] [6]

VOC gas leak detection using pyro-electric infrared sensors,

F. Erden, E. B. Soyer, B. U. Toreyin, and A. E. Cetin, “VOC gas leak detection using pyro-electric infrared sensors,” in Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference on . IEEE, 2010, pp. 1682–1685

work page 2010

[8] [7]

Method, device and system for determining the presence of volatile organic compounds (VOC) in video,

A. E. Cetin and B. U. Toreyin, “Method, device and system for determining the presence of volatile organic compounds (VOC) in video,” Apr. 30 2013, uS Patent 8,432,451

work page 2013