STORM: Strategic Orchestration of Modalities for Rare Event Classification

Ayan Banerjee; Payal Kamboj; Sandeep K.S. Gupta

REVIEW 5 major objections 6 minor 1 cited by

STORM: Strategic Orchestration of Modalities for Rare Event Classification

T0 review · 5 major / 6 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 STORM is an entropy-based algorithm that selects the minimal informative modality set for rare-event classification, and in seizure-onset-zone detection it shows the expert-derived temporal modality is redundant, with F1 unchanged at 91.6%.

desk verdict The entropy-imbalance selection criterion is asserted, not validated, and the single fragile evaluation makes the D2-redundancy finding interesting but unproven. read the letter →

arxiv 2412.02805 v1 pith:JTE6QR63 submitted 2024-12-03 cs.CV

Payal Kamboj , Ayan Banerjee , Sandeep K.S. Gupta This is my paper · ORCID

classification cs.CV

keywords modalityselectionrareeventclassificationentropyimbalanceseizureonsetzonemultimodallearningclassexpertknowledgers-fMRI

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

STORM is an entropy-based algorithm for deciding which data modalities a classifier actually needs when the target event is rare. It scores each candidate modality by the spread of class-wise entropies in its learned representation, on the theory that a modality with the smallest such spread has learned the most discriminative class structure. In a seizure-onset-zone detection study on 52 pediatric epilepsy patients, STORM identifies the raw spatial fMRI modality plus one expert-derived spatial feature set as sufficient; the expert-derived temporal modality is unnecessary. Omitting the temporal modality leaves accuracy at 84.6% and F1 at 91.6% unchanged, while omitting the spatial-derived features collapses F1 to 0%. The paper argues this replaces manual trial-and-error modality selection with a systematic, quantifiable criterion.

What carries the argument

The central object is the entropy imbalance metric $\eta_{M_d}$, computed by (i) taking a trained classifier's intermediate representation $F_{M_d}(y_i)$ for each instance, (ii) measuring each instance's local density via K-nearest-neighbor distances, (iii) converting those densities into a class-wise entropy $\theta_r$ using the entropy-based sampling formulas of [19], and (iv) taking the maximum deviation $\max_{c_r}(\theta_r - \mathbb{E}(\theta_r))$ across classes. STORM then uses the gain $\mathrm{EIG}(M_d) = \eta_R - \eta_{M_d}$ as a splitting criterion in a decision tree built with Hunt's algorithm, with thresholds on EIG equivalence, Gini impurity, and classifier confidence controlling when to stop or cascade classifiers.

What would settle it

Feed STORM a modality whose features are randomized within each class, so class labels carry no feature information; if the entropy-imbalance gain EIG for that modality stays high enough to be selected, the metric is not measuring discriminative content. A second check: on a dataset built so the rare class is visible only in the temporal modality, STORM should select that temporal modality, and if it drops it, the selection rule is wrong.

Watch

Extended reading notes

Core claim

The paper claims that modality selection for rare-event classification can be solved by measuring, for each candidate modality's learned representation, the spread of class-wise entropies: $\eta_{M_d} = \max_{c_r} (\theta_r^{M_d} - \mathbb{E}(\theta_r^{M_d}))$, where $\theta_r^{M_d}$ is the entropy of class $c_r$ under classifier $M_d$. A modality whose representation has the lowest entropy imbalance is taken to have learned the most discriminative class structure, and selection proceeds greedily via Hunt's algorithm using the entropy-imbalance gain $\mathrm{EIG}(M_d) = \eta_R - \eta_{M_d}$. In the seizure-onset-zone case study, applying STORM identifies the basic spatial modality plus the expert-derived spatial modality D1 as sufficient; the expert-derived temporal modality D2 is deemed unnecessary because excluding it leaves accuracy at 84.6% and F1 at 91.6% unchanged, while excluding D1 collapses F1 to 0%.

Load-bearing premise

The load-bearing premise is that the entropy-imbalance value $\eta_{M_d}$ of a modality's learned representation tracks how much discriminative information that modality actually carries for the rare class; if low entropy imbalance does not predict good rare-class separability, STORM's ranking of modalities is not grounded.

Editorial extensions

If this is right

On the 52-patient rs-fMRI dataset, the full modality set and the reduced set (basic plus D1) both reach 84.6% accuracy and 91.6% F1; dropping D2 costs nothing.
Removing the spatial-derived modality D1, or just its D1b rule, drops F1 to 0%, so at least one expert-derived spatial rule is load-bearing for SOZ detection.
Raw data alone is not enough: excluding all derived modalities drops F1 to 63.0%, and excluding the basic modality drops it to 67.0%, so basic and derived modalities contribute complementary information.
STORM gives a quantitative stopping rule for modality collection: a modality whose entropy-imbalance gain is within threshold of another's can be omitted before training a full multimodal model.
The approach is framed for rare-event biomedical tasks and is proposed for other expert-guided domains, such as coronary artery disease detection from ECG images versus extracted time-series leads.

Reading between the lines

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

By the same logic, the entropy-imbalance criterion should transfer to any multimodal pipeline with per-modality learned representations; a natural next experiment is the CAD case the paper sketches but does not run, comparing ECG images against extracted time-series leads.
The ablation pattern suggests feature-level selection is the next layer: since removing D1b alone collapses F1 to 0 exactly as removing all of D1 does, STORM could be run over individual expert rules rather than whole modalities, likely pruning D2a/D2b even within a selected modality.
A consequence of this design is a divergence between expert usefulness and machine usefulness: temporal features (D2) are reported as meaningful for human SOZ reading but contribute nothing to the classifier, implying that clinically meaningful modalities may still be algorithmically redundant, which matters for data collection and storage decisions.

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: the paper has a plausible idea and one potentially useful empirical observation, but the load-bearing metric is asserted rather than validated, and the single evaluation is too fragile to support the claims.

What's new: STORM names a specific way to combine class-wise entropy with a decision-tree search over modalities, and on this pediatric rs-fMRI dataset it finds that the temporal derived modality D2 adds nothing to SOZ detection. That last result is genuinely interesting if real, because it contradicts the expert assumption that both D1 and D2 matter.

Soft spots: the entropy imbalance ηMd is the core of the method, but the paper never shows that lower ηMd tracks better rare-class separability. The stress-test note is right: overlapping classes can have similar entropies, so ηMd≈0, while well-separated classes with unequal variances can have high ηMd. Without a synthetic test or a theoretical argument, the selection criterion isn't grounded. Second, Table I is hard to trust. No error bars or cross-validation on a dataset of 52 patients. The 0% rows for D1-excluded and D1b-excluded configurations look like pipeline failures, not measured performance—if removing one rule feature truly collapsed F1 from 91.6 to 0, that would suggest the model is not learning anything robust. Also, EIG is computed from trained classifiers on the same data used to select them, so there's a circularity that isn't addressed. The 'another case study' on CAD is a paragraph, not a case study, and there are no comparisons to existing modality-selection methods, including the ones they cite (e.g., He et al. [4]).

Who it's for: someone working on multi-modal medical ML who wants a concrete example of how not to validate a selection criterion. It could be useful as a cautionary case.

Recommendation: I'd send it to peer review because the underlying question is worthwhile and the empirical claim about D2 is testable. But the reviewers should ask for a validation of ηMd (synthetic experiments, comparison with feature-selection baselines) and a much more careful evaluation with confidence intervals. As it stands, I wouldn't cite it for the method.

Referee Report

5 major / 6 minor

Summary. This paper presents STORM, an entropy-based algorithm for selecting modalities in rare-event classification. The algorithm computes class-wise entropy imbalances of learned representations and selects modalities that maximize the entropy imbalance gain. The method is applied to seizure onset zone (SOZ) detection from resting-state fMRI, using a basic spatial modality and two expert-derived modalities (D1 spatial, D2 temporal). The reported results suggest that the basic modality and D1 are necessary, while D2 is unnecessary for SOZ classification. The paper also mentions a coronary artery disease case study but does not report results for it.

Significance. If validated, the STORM framework could offer a principled alternative to ad-hoc modality selection in biomedical AI, with the potential to reduce data collection and processing costs. The paper's strengths include its focus on a clinically relevant problem, the use of expert knowledge, and the presentation of a concrete algorithm with pseudocode. However, the central entropy-imbalance criterion is not theoretically derived or empirically validated, and the reported experiments lack error bars and statistical rigor. The contribution is therefore currently a proposal with a single case study, rather than an established method.

major comments (5)

[Section II.B, Eq. (5)] The assertion that the classifier with the lowest ηMd is ideally the best because 'representative class features were learned' is not justified. ηMd only measures the dispersion of class-wise entropies; it does not directly measure inter-class separation or the information content of a modality for discriminating the rare class. A task with overlapping classes can have low ηMd, while a task with well-separated classes but heterogeneous within-class densities can have high ηMd. The authors need to provide a controlled experiment (e.g., synthetic data with known discriminative structure) that demonstrates a monotonic relationship between ηMd and rare-class F1 or similar metrics.
[Section III.B, Table I] The 0% rows for 'Modality D1 excluded' and 'Modality D1b excluded' are not valid performance measurements; they indicate a degenerate classifier output. The authors should explain what occurred (e.g., no positive predictions) and report results over cross-validation folds with standard deviations. Additionally, the identical scores across 'All modalities included,' 'Modality D2 excluded,' 'Modality D2a excluded,' and 'Modality D2b excluded' suggest that D2 features are not being used in the final ensemble; the paper must verify that the pipeline correctly consumes D2 features and that the result is not an artifact of a code path.
[Section II.B, Eq. (6)] The entropy imbalance gain EIG compares ηR (from raw-data representations such as PSNR or VGG penultimate-layer features) with ηMd (from classifier intermediate representations). Since these representations have different dimensionalities and scales, the difference may reflect representation geometry rather than modality informativeness. The authors should either normalize the entropies or use comparable representation types when computing the gain.
[Section II.A] The class-wise entropy computation depends on the choice of distance metric and the nearest-neighbor count K, but the paper neither states the value of K nor analyzes the sensitivity of the results to K and to the thresholds ϵm, ϵg, and dth. These free parameters should be justified and a robustness analysis should be provided.
[Section III.C] The second case study (CAD detection) is described only as a potential application, with no experimental results. Since the abstract and introduction claim a general-purpose algorithm, the inclusion of a second case study would strengthen the empirical support; as it stands, the paper provides only a single clinical demonstration.

minor comments (6)

[Section I.C] The phrase 'Note than even though' should be 'Note that even though.'
[Section II.A] The use of PSNR as a representation for raw data is unclear, as PSNR is a scalar quality metric; the authors should clarify what distance between instances is being computed.
[Section II.B] The connection to Hunt's algorithm is mentioned but not explained; a brief description of how the decision tree is built would improve readability.
[Table I] The table notes N=52 but does not specify the number of SOZ instances; the class distribution should be reported to contextualize the rare-event setting.
[References] Reference [24] is incomplete ('R. U. et al.'); the full author list should be provided.
[Figure 2] Figure 2 is not described in the text; the caption should be self-contained and the figure should be referenced in the relevant section.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's central modality-selection result is supported by an empirical ablation, not by a definitional reduction or a self-citation chain.

full rationale

The paper's main conclusion—that the basic modality plus derived spatial modality D1 are sufficient while derived temporal modality D2 is unnecessary—is established by direct ablation experiments in Table I, where excluding D2 leaves F1 unchanged at 91.6% while excluding D1 collapses F1 to 0%. That is an empirical measurement, not a quantity forced by the paper's definitions. The entropy-imbalance metric ηMd in Eq. (5) and the gain EIG in Eq. (6) are heuristic criteria for ranking classifiers and modalities, but the paper does not substitute EIG for F1; rather, it reports actual classification performance as the evidence for D2's redundancy. The claim that a low ηMd implies that 'representative class features were learned' is asserted rather than proved, which is a validity or correctness risk, but it is not a circular step: no equation in the paper reduces the target result to the metric by construction. The expert-suggested rules D1a, D1b, D2a, and D2b are cited to prior works by overlapping authors, but those rules are presented as domain knowledge from clinical experts, not as mathematically derived premises whose validation is the present paper's conclusion. The self-citations are therefore not load-bearing in a way that makes the argument circular. The paper is self-contained against an external clinical dataset and standard classification metrics, so no significant circularity is found.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central selection criterion rests on an unproven assertion that entropy imbalance proxies discriminative power. The algorithm's behavior is governed by several hand-chosen thresholds that are not reported. The expert-derived features are self-cited, adding to the burden of demonstrating independent support.

free parameters (6)

K (number of nearest neighbors)
Used in Q(x_i) for density estimation; value not specified anywhere in the paper.
ϵm (entropy imbalance threshold)
Algorithm 1 uses ϵm to decide when classifiers are equivalent; value not given.
ϵg (Gini impurity threshold)
Algorithm 1 stops when Gini index < ϵg; value not given.
dth (dependability threshold)
Used to break ties between classifiers in Algorithm 1; value not given.
SVM SOZ score cutoff = 0.9
Used in final label fusion (Section II.B); chosen by hand, not derived.
SMOTE parameters
SMOTE applied before SVM; number of neighbors and other settings not reported.

assumptions (5)

ad hoc to paper Low entropy imbalance ηMd is a valid proxy for a modality's discriminative value for the rare class.
Stated in Section II.B: 'Ideally the best classifier should have an intermediate representation that has the lowest value of ηMd'. No proof or validation is provided.
domain assumption Class-wise entropy (from Li et al. [19]) is an appropriate measure of class imbalance and modality informativeness.
Adopted as the foundation of the assessment step (Section II.A).
domain assumption Expert rules D1a, D1b, D2a, D2b correctly capture SOZ characteristics.
Taken from the authors' prior work [11], [12] (Section I.C).
domain assumption Euclidean distance is appropriate for all representation spaces (PSNR, CNN features, expert-derived features).
Section II.A states 'The distance metric dist(.,.) for all cases is Euclidean distance.'
domain assumption ICA decomposition yields meaningful spatial and temporal modalities with SOZ prevalence below 5%.
Section I.C states ICA typically produces 100-200 ICs with less than 5% SOZ.

how reviews work

0 comments

Cite this review

Pith. "Pith review of STORM: Strategic Orchestration of Modalities for Rare Event Classification." pith.science (2026). https://pith.science/paper/JTE6QR63

@misc{pith2026241202805,
  author       = {Pith},
  title        = {Pith review of: STORM: Strategic Orchestration of Modalities for Rare Event Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JTE6QR63}},
  note         = {Machine review of arXiv:2412.02805}
}

read the original abstract

In domains such as biomedical, expert insights are crucial for selecting the most informative modalities for artificial intelligence (AI) methodologies. However, using all available modalities poses challenges, particularly in determining the impact of each modality on performance and optimizing their combinations for accurate classification. Traditional approaches resort to manual trial and error methods, lacking systematic frameworks for discerning the most relevant modalities. Moreover, although multi-modal learning enables the integration of information from diverse sources, utilizing all available modalities is often impractical and unnecessary. To address this, we introduce an entropy-based algorithm STORM to solve the modality selection problem for rare event. This algorithm systematically evaluates the information content of individual modalities and their combinations, identifying the most discriminative features essential for rare class classification tasks. Through seizure onset zone detection case study, we demonstrate the efficacy of our algorithm in enhancing classification performance. By selecting useful subset of modalities, our approach paves the way for more efficient AI-driven biomedical analyses, thereby advancing disease diagnosis in clinical settings.

Figures

Figures reproduced from arXiv: 2412.02805 by the authors.

**Figure 2.** SOZ classification using STORM. Step 1. If there is a tie between classifiers, then the classifier with confidence score > dth is used to compute EIG(Md). Stopping condition: The process continues until the training set is exhausted or validation accuracy remains unchanged for consecutive cycles. Algorithm 1 STORM Algorithm Input: Raw data Y , Rare class cr, Thresholds ϵm, ϵg, Dependability threshold dth, set of cla… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Generating customized prompts for Zero-Shot Rare Event Medical Image Classification using LLM
cs.CV 2025-01 reject novelty 3.0 of 10

CuKPL converts medical images into text descriptions of expert-defined features and asks GPT-4o to classify them as seizure onset zone or not, claiming zero-shot performance that is actually worse than supervised deep...

Reference graph

Works this paper leans on

24 extracted references · 16 canonical work pages · cited by 1 Pith paper

[4]

Efficient modality selection in multimodal learning,

Y . He, R. Cheng, G. Balasubramaniam, Y .-H. H. Tsai, and H. Zhao, “Efficient modality selection in multimodal learning,” Journal of Machine Learning Research, vol. 25, no. 47, pp. 1–39, 2024

work page 2024
[1]

What makes multi-modal learning better than single (provably),

Y . Huang, C. Du, Z. Xue, X. Chen, H. Zhao, and L. Huang, “What makes multi-modal learning better than single (provably),” in Advances in Neural Information Processing Systems (M. Ranzato, A. Beygelzimer, Y . Dauphin, P. Liang, and J. W. Vaughan, eds.), vol. 34, pp. 10944–10956, Curran Associates, Inc., 2021

work page 2021
[2]

Transfer: Cross modality knowledge transfer using adversarial networks–a study on gesture recognition,

P. Kamboj, A. Banerjee, and S. K. Gupta, “Transfer: Cross modality knowledge transfer using adversarial networks–a study on gesture recognition,” arXiv preprint arXiv:2306.15114 , 2023

arXiv 2023
[3]

Multimodal deep networks for text and image-based document classification,

N. Audebert, C. Herold, K. Slimani, and C. Vidal, “Multimodal deep networks for text and image-based document classification,” CoRR, vol. abs/1907.06370, 2019

arXiv 1907
[5]

mslam: Massively multilingual joint pretraining for speech and text,

A. Bapna, C. Cherry, Y . Zhang, Y . Jia, M. Johnson, Y . Cheng, S. Khanuja, J. Riesa, and A. Conneau, “mslam: Massively multilingual joint pretraining for speech and text,” arXiv preprint arXiv:2202.01374 , 2022

arXiv 2022
[6]

Framework for developing and evaluating ethical collaboration between expert and machine,

A. Banerjee, P. Kamboj, and S. Gupta, “Framework for developing and evaluating ethical collaboration between expert and machine,” arXiv preprint arXiv:2411.10983, 2024

arXiv 2024
[7]

Sensor and sensor fusion technology in autonomous vehicles: A review,

D. J. Yeong, G. Velasco-Hernandez, J. Barry, and J. Walsh, “Sensor and sensor fusion technology in autonomous vehicles: A review,” Sensors, vol. 21, no. 6, p. 2140, 2021

2021
[8]

High fidelity fast simulation of human in the loop human in the plant (hil-hip) systems,

A. Banerjee, P. Kamboj, A. Maity, R. Salian, and S. Gupta, “High fidelity fast simulation of human in the loop human in the plant (hil-hip) systems,” in Proceedings of the Int’l ACM Conference on Modeling Analysis and Simulation of Wireless and Mobile Systems , MSWiM ’23, (New York, NY , USA), p. 199–203, Association for Computing Machinery, 2023

work page 2023

Show all 24 references

[9]

Cps-llm: Large language model based safe usage plan generator for human- in-the-loop human-in-the-plant cyber-physical system,

A. Banerjee, A. Maity, P. Kamboj, and S. K. Gupta, “Cps-llm: Large language model based safe usage plan generator for human- in-the-loop human-in-the-plant cyber-physical system,” arXiv preprint arXiv:2405.11458, 2024

2024 arXiv
[10]

Use of multi-modal data and machine learning to improve cardiovascular disease care,

S. Amal, L. Safarnejad, J. Omiye, I. Ghanzouri, J. Cabot, and E. Ross, “Use of multi-modal data and machine learning to improve cardiovascular disease care,” Frontiers in Cardiovascular Medicine , vol. 9, p. 840262, 2022

2022
[11]

The Expert’s Knowledge combined with AI outperforms AI Alone in Seizure Onset Zone Localization using resting state fMRI,

P. Kamboj, A. Banerjee, V . L. Boerwinkle, and S. K. S. Gupta, “The Expert’s Knowledge combined with AI outperforms AI Alone in Seizure Onset Zone Localization using resting state fMRI,”Frontiers in Neurology, vol. 14, 2024

2024
[12]

Automated seizure onset zone locator from resting-state functional MRI in drug-resistant epilepsy,

A. Banerjee, P. Kamboj, S. N. Wyckoff, B. L. Sussman, S. K. S. Gupta, and V . L. Boerwinkle, “Automated seizure onset zone locator from resting-state functional MRI in drug-resistant epilepsy,” Frontiers in Neuroimaging, vol. 1, p. 1007668, 2023

2023
[13]

Merging deep learning with expert knowledge for seizure onset zone localization from rs-fmri in pediatric pharmaco resistant epilepsy,

P. Kamboj, A. Banerjee, S. K. Gupta, and V . L. Boerwinkle, “Merging deep learning with expert knowledge for seizure onset zone localization from rs-fmri in pediatric pharmaco resistant epilepsy,” arXiv preprint arXiv:2306.05572, 2023

2023 arXiv
[14]

Multimodal machine learning in precision health: A scoping review,

A. Kline, H. Wang, Y . Li, et al. , “Multimodal machine learning in precision health: A scoping review,” npj Digital Medicine , vol. 5, p. 171, 2022

2022
[15]

Diagnostic performance of electrocardiography in the assessment of significant coronary artery disease and its anatomical size in comparison with coronary angiography,

S. Mahmoodzadeh, M. Moazenzadeh, H. Rashidinejad, and M. Sheikhvatan, “Diagnostic performance of electrocardiography in the assessment of significant coronary artery disease and its anatomical size in comparison with coronary angiography,” Journal of Research in Medical Scienc...

2011
[16]

Expert knowledge driven human-ai collaboration for medical imaging: a study on epileptic seizure onset zone identification,

A. Banerjee, P. Kamboj, and S. K. S. Gupta, “Expert knowledge driven human-ai collaboration for medical imaging: a study on epileptic seizure onset zone identification,” Transactions on Artificial Intelligence , 2024

2024
[17]

Edgcon: auto-assigner of iconicity ratings grounded by lexical properties to aid in generation of technical gestures,

S. Hossain, P. Kamboj, A. Maity, T. Azuma, A. Banerjee, and S. Gupta, “Edgcon: auto-assigner of iconicity ratings grounded by lexical properties to aid in generation of technical gestures,” in Proceedings of the 38th ACM/SIGAPP Symposium on Applied Computing , pp. 3–10, 2023

2023
[18]

Evaluation of rare event detection,

M. Sokolova, K. El Emam, S. Chowdhury, E. Neri, S. Rose, and E. Jonker, “Evaluation of rare event detection,” in Advances in Artificial Intelligence: 23rd Canadian Conference on Artificial Intelligence, Canadian AI, Proceedings 23, pp. 379–383, Springer, 2010

2010
[19]

Entropy-based sampling approaches for multi- class imbalanced problems,

L. Li, H. He, and J. Li, “Entropy-based sampling approaches for multi- class imbalanced problems,” IEEE Transactions on Knowledge and Data Engineering, vol. 32, no. 11, pp. 2159–2170, 2020

2020
[20]

Deep facial analysis: A new phase I epilepsy evaluation using computer vision,

D. Ahmedt-Aristizabal, C. Fookes, K. Nguyen, S. Denman, S. Sridharan, and S. Dionisio, “Deep facial analysis: A new phase I epilepsy evaluation using computer vision,” Epilepsy & Behavior , vol. 82, pp. 17–24, 2018

2018
[21]

More Interpretable Decision Trees,

E. Gilmore, V . Estivill-Castro, and R. Hexel, “More Interpretable Decision Trees,” in Hybrid Artificial Intelligent Systems (H. Sanjurjo González, I. Pastor López, P. García Bringas, H. Quintián, and E. Corchado, eds.), (Cham), pp. 280–292, Springer International Publishing, 2021

2021
[22]

A simple method for measuring inequality,

T. Sitthiyot and K. Holasut, “A simple method for measuring inequality,” Palgrave Communications, vol. 6, 2020

2020
[23]

SMOTE: synthetic minority over-sampling technique,

K. W. Bowyer, N. V . Chawla, L. O. Hall, and W. P. Kegelmeyer, “SMOTE: synthetic minority over-sampling technique,” CoRR, vol. abs/1106.1813, 2011

2011 arXiv
[24]

Automated echocardiographic detection of severe coronary artery disease using artificial intelligence,

R. U. et al., “Automated echocardiographic detection of severe coronary artery disease using artificial intelligence,” Cardiovascular Imaging , vol. 15, no. 5, p. s.l., 2022

2022

Pith tools

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

[1] [4]

Efficient modality selection in multimodal learning,

Y . He, R. Cheng, G. Balasubramaniam, Y .-H. H. Tsai, and H. Zhao, “Efficient modality selection in multimodal learning,” Journal of Machine Learning Research, vol. 25, no. 47, pp. 1–39, 2024

work page 2024

[2] [1]

What makes multi-modal learning better than single (provably),

Y . Huang, C. Du, Z. Xue, X. Chen, H. Zhao, and L. Huang, “What makes multi-modal learning better than single (provably),” in Advances in Neural Information Processing Systems (M. Ranzato, A. Beygelzimer, Y . Dauphin, P. Liang, and J. W. Vaughan, eds.), vol. 34, pp. 10944–10956, Curran Associates, Inc., 2021

work page 2021

[3] [2]

Transfer: Cross modality knowledge transfer using adversarial networks–a study on gesture recognition,

P. Kamboj, A. Banerjee, and S. K. Gupta, “Transfer: Cross modality knowledge transfer using adversarial networks–a study on gesture recognition,” arXiv preprint arXiv:2306.15114 , 2023

arXiv 2023

[4] [3]

Multimodal deep networks for text and image-based document classification,

N. Audebert, C. Herold, K. Slimani, and C. Vidal, “Multimodal deep networks for text and image-based document classification,” CoRR, vol. abs/1907.06370, 2019

arXiv 1907

[5] [5]

mslam: Massively multilingual joint pretraining for speech and text,

A. Bapna, C. Cherry, Y . Zhang, Y . Jia, M. Johnson, Y . Cheng, S. Khanuja, J. Riesa, and A. Conneau, “mslam: Massively multilingual joint pretraining for speech and text,” arXiv preprint arXiv:2202.01374 , 2022

arXiv 2022

[6] [6]

Framework for developing and evaluating ethical collaboration between expert and machine,

A. Banerjee, P. Kamboj, and S. Gupta, “Framework for developing and evaluating ethical collaboration between expert and machine,” arXiv preprint arXiv:2411.10983, 2024

arXiv 2024

[7] [7]

Sensor and sensor fusion technology in autonomous vehicles: A review,

D. J. Yeong, G. Velasco-Hernandez, J. Barry, and J. Walsh, “Sensor and sensor fusion technology in autonomous vehicles: A review,” Sensors, vol. 21, no. 6, p. 2140, 2021

2021

[8] [8]

High fidelity fast simulation of human in the loop human in the plant (hil-hip) systems,

A. Banerjee, P. Kamboj, A. Maity, R. Salian, and S. Gupta, “High fidelity fast simulation of human in the loop human in the plant (hil-hip) systems,” in Proceedings of the Int’l ACM Conference on Modeling Analysis and Simulation of Wireless and Mobile Systems , MSWiM ’23, (New York, NY , USA), p. 199–203, Association for Computing Machinery, 2023

work page 2023