Pith. sign in

REVIEW 3 major objections 5 minor 32 references

NSF-MAP: Neurosymbolic Multimodal Fusion for Robust and Interpretable Anomaly Prediction in Assembly Pipelines

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims that fusing time-series sensor data with camera images, through decision-level fusion enhanced by transfer learning and ontology-based knowledge infusion, predicts assembly-line anomalies with 93% weighted F1, far above…

desk verdict The knowledge-infusion penalty in Eq. 8 is non-differentiable and cannot explain the P3 improvement under Adam—a load-bearing flaw that needs a mechanism or a rerun, despite a solid fusion baseline and useful dataset. read the letter →

arxiv 2505.06333 v1 pith:CMXBQGCI submitted 2025-05-09 cs.LG cs.AI

Chathurangi Shyalika , Renjith Prasad , Fadi El Kalach , Revathy Venkataramanan , Ramtin Zand , Ramy Harik , Amit Sheth This is my paper · ORCID
classification cs.LGcs.AI
keywords anomalypredictionmultimodalfusionneurosymbolicAItransferlearningknowledgeinfusionassemblypipelinetimeseriesimageclassification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that neurosymbolic fusion of multivariate sensor time series and camera images can predict the next step's anomaly type in a rocket assembly pipeline, and that each of three additions—decision-level fusion, transfer learning (freezing the encoder), and knowledge-infused penalties from a process ontology—contributes to the gain. On an 80/20 cycle-wise split the full model reaches 93% weighted F1 and accuracy versus 63% for the time-series-only baseline and 72% for plain fusion. The authors also contribute two derived datasets (multimodal and analog) and an ontology-based explanation layer that ties predicted sensor values to expected ranges per assembly state. The claim is significant because multimodal anomaly prediction in manufacturing lacks public datasets and interpretable, robust models.

What carries the argument

The load-bearing object is the decision-level fusion vector $z = [f_V; h_T]$, where $f_V$ comes from a pretrained EfficientNet-B0 with classification layer removed, applied to YOLO-FF-cropped region-of-interest images from cycle states 4 and 9, and $h_T$ is the reconstruction output of a time-series autoencoder whose encoder is frozen during transfer-learning training. A fully connected network maps $z$ to predicted next-step sensor values and anomaly label, and a weighted MSE loss with a knowledge-infusion penalty term $P$ adds a cost whenever predictions fall outside the ontology's expected sensor ranges for the current cycle state. The Dynamic Process Ontology over the 21 cycle states is what makes the system neurosymbolic: it encodes sensor-robot-state relationships and state-dependent expected ranges used both in the loss and in user-level explanations.

What would settle it

Re-annotate a random sample of the test cycles using independent physical inspection (e.g., weighing the rocket or post-assembly verification) instead of the expert mapping function, and recompute the model's weighted F1 against this independent ground truth; if the fusion model no longer beats the time-series-only baseline, the reported gains are an artifact of the label mapping.

Watch

Extended reading notes

Core claim

Fusing image embeddings from a pretrained EfficientNet-B0 with the latent reconstruction of a time-series autoencoder, then feeding the concatenated vector to a fully connected predictor, yields a 72% accurate anomaly-type predictor on the 80/20 split; freezing the autoencoder encoder (transfer learning) raises this to 88%; adding a penalty when predicted sensor values contradict the process ontology's expected ranges raises it to 93% weighted F1 and accuracy on the same split. The model predicts the next time step's sensor values and anomaly label, and the ontology doubles as an explainability layer that states which variable, robot function, and expected range explain an anomaly, and can flag model misclassifications (e.g., predicting a NoNose anomaly in a state where that anomaly cannot occur). On the 60/40 split the full model still leads with 83% weighted F1 and accuracy.

Load-bearing premise

The ground-truth anomaly labels, the 21 cycle states, and the filtering of images to states 4 and 9 all come from a mapping function supplied by domain experts; if that mapping is inaccurate, the model's high accuracy reflects consistently mislabeled data rather than genuine detection of assembly anomalies.

Editorial extensions

If this is right

  • If correct, the 93% weighted F1 on this testbed means the model is deployable: it can flag missing rocket body and nose anomalies in real time from OPC-UA sensor feeds and camera streams.
  • The ablation isolates three separable gains: fusion over time-series alone, transfer learning over plain fusion, and knowledge infusion over transfer learning, so future work can build on any one component independently.
  • The ontology's ability to catch impossible predictions (e.g., an anomaly type that does not exist in the current cycle state) offers a model-independent safety net for production.
  • The derived multimodal and analog datasets and the process ontology are released publicly, so other groups can benchmark new fusion or knowledge-infusion methods against this pipeline.

Reading between the lines

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

  • The reported gain from freezing the encoder is surprisingly large and may partly come from regularizing the autoencoder rather than from transfer learning per se; an ablated comparison of freezing versus weight decay or dropout would isolate which mechanism truly drives the improvement.
  • If the expert mapping between sensor records and cycle states contains systematic error, the knowledge-infusion penalty could reinforce those errors, and the gain from the full model should be re-measured on independently labeled anomalies.
  • The image stream only enters at states 4 and 9; the authors state other states use time series alone, so a multi-state extension that uses images at all states or learns when to attend to images could generalize the gain beyond two states.
  • The ontology's expected sensor ranges are static bounds per state; dynamic ranges fit from historical data would make the penalty adaptive and could be tested by comparing fixed versus learned ranges.
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.

Referee Report

3 major / 5 minor

Summary. The paper proposes NSF-MAP, a decision-level fusion framework that combines time-series autoencoder features with ImageNet-pretrained EfficientNet image features to predict the next sensor values and anomaly label in an assembly-pipeline dataset. Three variants are compared: P1 (plain fusion), P2 (fusion plus transfer learning, with the time-series encoder frozen), and P3 (P2 plus a knowledge-infusion penalty derived from a Dynamic Process Ontology). The authors report that P3 reaches 93% weighted F1/accuracy on an 80/20 cycle-wise split, versus 63% for the time-series baseline and 72% for plain fusion, and they also describe an ontology-based explainability module and a real-time deployment at the Future Factories testbed.

Significance. If the reported results hold, the paper would provide a useful blueprint for multimodal anomaly prediction in manufacturing, and the public release of the preprocessed multimodal dataset, code, and deployment demo is a genuine service to the community. The ontology-based explainability mechanism is a valuable contribution independent of the accuracy claims. However, the central novelty—the knowledge-infusion penalty that distinguishes P3 from P2—is not supported by the experiments as described, because the penalty term is non-differentiable and cannot influence Adam-based training. In addition, the table compares the proposed prediction models against an image-only detection model that solves a different task, and that model outperforms the proposed method. These issues are load-bearing for the paper's main claim that neurosymbolic knowledge infusion improves fusion performance.

major comments (3)
  1. [§5, Eq. (8) and Eq. (12); §6.1 P3] The knowledge-infusion penalty P in Eq. (8) is a sum of indicator functions that are piecewise constant in the model output, so their gradient with respect to the model parameters is zero almost everywhere. Under the stated Adam optimization in Section 6, adding λP to the WMSE loss in Eq. (12) therefore does not change the parameter updates for almost all mini-batches. Yet Table 2 reports that P3 (DLF+TL+KIL) improves over P2 by roughly 3–5 points in F1 and accuracy. The manuscript describes no surrogate gradient, straight-through estimator, penalty-based model selection, or augmented Lagrangian that would allow the indicator penalty to affect training. This is a load-bearing gap: the paper's central neurosymbolic contribution is the knowledge-infused penalty, and the reported P3 gain cannot be explained by the mechanism as specified. Please either provide the missing mechanism or re-run the experiments with a differentiable penalty (e.g., a soft hinge or logistic penalty) and report whether the P3 advantage persists.
  2. [§7, Table 2 and Figure 3] The image-only baseline B2 is described in Section 6 as an EfficientNet-B0 trained for anomaly detection on preprocessed images, i.e., a current-frame classification model, while all other methods predict the next time step's anomaly label from current time-series and image data. Placing B2 in the same results table and using it as a comparison baseline is an apples-to-oranges comparison: B2 has access to the current image's ground-truth content and does not perform prediction. This is especially problematic because B2 achieves 97% F1/accuracy in the 80/20 split, higher than the proposed P3 (93%). The evaluation should either remove B2 from the prediction comparison or reformulate it as a prediction model (e.g., by training it on time-shifted labels and evaluating on the same future-step task), and the text should clearly state that B2 is a detection-only reference, not a prediction baseline.
  3. [§3, Data Preparation; §5, Process Ontology] The paper states that cycle-state information, anomaly types, and the mapping from sensor records to ground-truth labels were obtained from a mapping function supplied by domain experts, and that images were filtered to states 4 and 9 using domain expertise. The ontology's expected sensor ranges also depend on this domain knowledge. All reported accuracies are therefore conditioned on the correctness and consistency of this mapping, but the manuscript provides no validation of the mapping (e.g., comparison against an independent audit, inter-annotator agreement, or a sensitivity analysis showing that results are stable under plausible mapping errors). Please add such an analysis or explicitly state the limitations this imposes on the reported numbers.
minor comments (5)
  1. [§5, Eq. (10) and Eq. (12)] The notation in Eq. (12) is inconsistent with Eq. (10): Eq. (10) minimizes over n, j ∈ {4,9}, and k, while Eq. (12) introduces new sums over c, s, and k with indices i, j, l that do not match the preceding definitions. Please harmonize the notation and define all indices and ranges.
  2. [§7, Table 2] The table header contains the typo "F1-Socre" and the column order changes between the two splits; please check that the column names match the reported metrics.
  3. [§6.1] The hyperparameter values for λ in Eq. (12), the WMSE class weights w, and the autoencoder hidden/latent dimensions for P2/P3 are not reported; these are needed to reproduce the experiments.
  4. [§6, Experimental Setup] The paper reports mean ± standard deviation but does not state the number of independent runs or the random-seed protocol; please specify how the means and standard deviations were computed.
  5. [§8, Deployment] There are several bracket typos in Section 8 and the appendix, e.g., "See deployment setup in Appendix 10.3]" and "explained in Appendix 10.3]"; these should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: fusion, transfer learning, and the knowledge penalty are evaluated on held-out cycles with external ImageNet features and domain-specified ranges, not on quantities fitted to the target labels.

full rationale

The NSF-MAP derivation chain is self-contained for the purposes of the circularity audit. The model predicts next-step sensor values and anomaly labels from current-step time series and images (Eqs. 2-7, 10-12), with held-out cycle-wise splits reported in Table 2; the target is the next time step, not a refit of the input. The transfer-learning variant freezes the time-series encoder and uses an ImageNet-pretrained EfficientNet (Eq. 1, Section 6), an external backbone that does not encode the paper's own anomaly labels. The knowledge-infusion penalty (Eq. 8) is defined using sensor ranges taken from the Process Ontology, described as domain knowledge gathered from the process ontology rather than fitted to the training labels, so it does not reduce to the supervised signal by construction. The ground-truth anomaly labels are extracted using a mapping function provided by domain experts (Section 3), and the ontology ranges are likewise expert-specified; this is external knowledge, not a term derived from the model output. Self-citations to the FF dataset (Harik et al., 2024) and to AssemAI/YOLO-FF (Prasad et al., 2024) supply input data and bounding boxes; these are preconditions of the experiment, not conclusions that the paper's equations force. Appendix 10.3 flags deployment drift and synchronization issues, but these are limitations about real-world transfer, not circular derivations. The reported non-differentiability of the indicator penalty (Eq. 8) under Adam is a correctness risk about whether P3's gain can be explained by the stated loss, not a circularity of the kind where a prediction equals its input by construction; per the audit rules, that concern belongs in correctness review, not in the circularity score.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The paper introduces a Dynamic Process Ontology as a knowledge source, but it is a software artifact, not a physical entity in the sense of a new particle or force. The free parameters are the penalty weight, class weights, architecture sizes, and the choice of filtered cycle states. The axioms are domain assumptions about label correctness, image visibility, and ontology range accuracy.

free parameters (4)
  • penalty weight lambda = not stated
    Hyperparameter λ in Eq. 12 controls the strength of the knowledge-infusion penalty; the paper does not report its value or tuning procedure.
  • WMSE class weights = not stated
    Weights w in Eq. 9 are claimed to address class imbalance, but the paper does not state how they are computed (e.g., inverse frequency or a learned weighting).
  • autoencoder hidden and latent sizes = hidden 64, latent 128
    Architecture choices for the time-series autoencoder (Section 6.1) that affect capacity and are not derived from data.
  • filtered cycle states = states 4 and 9
    Only images from cycle states 4 and 9 are used for fusion, based on domain expertise; this choice defines the fusion baseline and the image branch contribution.
assumptions (3)
  • domain assumption The domain-expert mapping function assigning cycle states and anomaly labels to raw records is correct.
    All supervised labels and the filtering of images to states 4 and 9 rely on this mapping (Section 3).
  • domain assumption Rocket parts are visible to cameras only in cycle states 4 and 9, so images from other states are uninformative.
    Stated in Section 3: images from cycle four and part of cycle nine are filtered to focus on the region of interest.
  • domain assumption The sensor ranges R in the Process Ontology are accurate for the dataset and the deployment environment.
    The knowledge-infusion penalty (Eq. 8) depends on these ranges; Appendix 10.3 notes that data collected eight months prior may have different sensor ranges.

how reviews work

0 comments
Cite this review

Pith. "Pith review of NSF-MAP: Neurosymbolic Multimodal Fusion for Robust and Interpretable Anomaly Prediction in Assembly Pipelines." pith.science (2026). https://pith.science/paper/CMXBQGCI

@misc{pith2026250506333,
  author       = {Pith},
  title        = {Pith review of: NSF-MAP: Neurosymbolic Multimodal Fusion for Robust and Interpretable Anomaly Prediction in Assembly Pipelines},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CMXBQGCI}},
  note         = {Machine review of arXiv:2505.06333}
}
read the original abstract

In modern assembly pipelines, identifying anomalies is crucial in ensuring product quality and operational efficiency. Conventional single-modality methods fail to capture the intricate relationships required for precise anomaly prediction in complex predictive environments with abundant data and multiple modalities. This paper proposes a neurosymbolic AI and fusion-based approach for multimodal anomaly prediction in assembly pipelines. We introduce a time series and image-based fusion model that leverages decision-level fusion techniques. Our research builds upon three primary novel approaches in multimodal learning: time series and image-based decision-level fusion modeling, transfer learning for fusion, and knowledge-infused learning. We evaluate the novel method using our derived and publicly available multimodal dataset and conduct comprehensive ablation studies to assess the impact of our preprocessing techniques and fusion model compared to traditional baselines. The results demonstrate that a neurosymbolic AI-based fusion approach that uses transfer learning can effectively harness the complementary strengths of time series and image data, offering a robust and interpretable approach for anomaly prediction in assembly pipelines with enhanced performance. \noindent The datasets, codes to reproduce the results, supplementary materials, and demo are available at https://github.com/ChathurangiShyalika/NSF-MAP.

Figures

Figures reproduced from arXiv: 2505.06333 by the authors.

Figure 1
Figure 1. Architecture of NSF-MAP: Integration of time series and images for anomaly prediction, involving preprocessing, feature extraction [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Experimental Results of Predicting Differ [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Performance of the Proposed Approach and Baselines with Varied Training and Testing Splits and knowledge-infused learning approach gives the best re￾sults in detecting five out of seven types of classes [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 6
Figure 6. Figure 6: A Snapshot of Dynamic Process Ontology showcasing the [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 4
Figure 4. Figure 4: A Rocket Assembled by the Future Factories Lab. Any [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 7
Figure 7. Figure 7: illustrates the architecture for real-time deployment of the NSF-MAP model and process ontology onto the man￾ufacturing system. It highlights integrating the trained model with the OPC-UA server for data retrieval and the connection to cameras for image acquisition, en…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 29 canonical work pages

  1. [1]

    Neuro-symbolic empow- ered denoising diffusion probabilistic models for real-time anomaly detection in industry 4.0: Wild-and-crazy-idea paper

    [Capogrosso et al., 2023] Luigi Capogrosso, Alessio Mas- colini, Federico Girella, Geri Skenderi, Sebastiano Ga- iardelli, Nicola Dall’Ora, Francesco Ponzio, Enrico Frac- caroli, Santa Di Cataldo, Sara Vinco, Enrico Macii, Franco Fummi, and Marco Cristani. Neuro-symbolic empow- ered denoising diffusion probabilistic models for real-time anomaly detection ...

    work page 2023

  2. [6]

    A framework for anomaly classification using deep transfer learning approach

    [Jayaswal and Dixit, 2021] Ruchi Jayaswal and Manish Dixit. A framework for anomaly classification using deep transfer learning approach. Revue d’Intelligence Artifi- cielle, 35(3),

    work page 2021

  3. [9]

    [Lughofer et al., 2022] Edwin Lughofer, Patrick Zorn, and Edmund Marth. Transfer learning of fuzzy classifiers for optimized joint representation of simulated and measured data in anomaly detection of motor phase currents.Applied Soft Computing, 124:109013,

    work page 2022

  4. [11]

    Real-time rocket assembly line manufacturing datasets

    [McCormick et al., 2025] MR McCormick, Fadi El Kalach, Mojtaba A Farahani, Ramy Harik, and Thorsten Wuest. Real-time rocket assembly line manufacturing datasets

    work page 2025

  5. [14]

    Anomaly detection from system tracing data using multimodal deep learning

    [Nedelkoski et al., 2019] Sasho Nedelkoski, Jorge Cardoso, and Odej Kao. Anomaly detection from system tracing data using multimodal deep learning. 2019 IEEE 12th International Conference on Cloud Computing (CLOUD), pages 179–186,

    work page 2019

  6. [16]

    Assemai: Inter- pretable image-based anomaly detection for manufactur- ing pipelines

    [Prasad et al., 2024] Renjith Prasad, Chathurangi Shyalika, Fadi El Kalach, Revathy Venkataramanan, Ramtin Zand, Ramy Harik, and Amit Sheth. Assemai: Inter- pretable image-based anomaly detection for manufactur- ing pipelines. In 2024 International Conference on Ma- chine Learning and Applications (ICMLA) , pages 1720– 1727,

    work page 2024

  7. [17]

    Mfgan: multimodal fusion for industrial anomaly detection using attention- based autoencoder and generative adversarial network

    [Qu et al., 2024] Xinji Qu, Zhuo Liu, Chase Q Wu, Aiqin Hou, Xiaoyan Yin, and Zhulian Chen. Mfgan: multimodal fusion for industrial anomaly detection using attention- based autoencoder and generative adversarial network. Sensors, 24(2):637,

    work page 2024

  8. [18]

    Signal processing technologies for activity-aware smart textiles

    [Roggen et al., 2013] Daniel Roggen, Gerhard Tr ¨oster, and Andreas Bulling. Signal processing technologies for activity-aware smart textiles. In Multidisciplinary know- how for smart-textiles developers , pages 329–365. Else- vier,

    work page 2013

Show all 32 references
  1. [19]

    Fusion of electronic health records and radiographic images for a multimodal deep learning prediction model of atypical femur fractures

    [Schilcher et al., 2024] J¨org Schilcher, Alva Nilsson, Oliver Andlid, and Anders Eklund. Fusion of electronic health records and radiographic images for a multimodal deep learning prediction model of atypical femur fractures. Computers in Biology and Medicine, 168:107704,

    2024

  2. [20]

    Design of an iterative method for enhanced multimodal time series analysis using graph attention networks, variational graph autoencoders, and transfer learning

    [Series, 2024] Enhanced Multimodal Time Series. Design of an iterative method for enhanced multimodal time series analysis using graph attention networks, variational graph autoencoders, and transfer learning. J. Electrical Systems, 20(5s):2579–2598,

    2024

  3. [21]

    Sheth, Kaushik Roy, and Manas Gaur

    [Sheth et al., 2023] Ami N. Sheth, Kaushik Roy, and Manas Gaur. Neurosymbolic ai - why, what, and how. ArXiv, abs/2305.00813,

    2023 arXiv

  4. [22]

    Tan and Q

    [Tan and Le, 2019] M. Tan and Q. V . Le. Efficientnet: Re- thinking model scaling for convolutional neural networks. Proceedings of the International Conference on Machine Learning (ICML), 97:6105–6114,

    2019

  5. [23]

    Transfer learning

    [Torrey and Shavlik, 2010] Lisa Torrey and Jude Shavlik. Transfer learning. In Handbook of research on machine learning applications and trends: algorithms, methods, and techniques, pages 242–264. IGI global,

    2010

  6. [26]

    A comprehensive survey of deep transfer learning for anomaly detection in industrial time series: Methods, applications, and direc- tions

    [Yan et al., 2024] Peng Yan, Ahmed Abdulkadir, Paul- Philipp Luley, Matthias Rosenthal, Gerrit A Schatte, Ben- jamin F Grewe, and Thilo Stadelmann. A comprehensive survey of deep transfer learning for anomaly detection in industrial time series: Methods, applications, and dire...

    2024

  7. [27]

    A robust hybrid deep learning model for spatiotemporal image fusion

    [Yang et al., 2021] Zijun Yang, Chunyuan Diao, and Bo Li. A robust hybrid deep learning model for spatiotemporal image fusion. Remote Sensing, 13(24):5005,

    2021

  8. [28]

    Efficient kpi anomaly detection through transfer learning for large- scale web services

    [Zhang et al., 2022] Shenglin Zhang, Zhenyu Zhong, Dong- wen Li, Qiliang Fan, Yongqian Sun, Man Zhu, Yuzhi Zhang, Dan Pei, Jiyan Sun, Yinlong Liu, et al. Efficient kpi anomaly detection through transfer learning for large- scale web services. IEEE Journal on Selected Areas in ...

    2022

  9. [29]

    Multimodal fusion on low-quality data: A comprehensive survey

    [Zhang et al., 2024] Qingyang Zhang, Yake Wei, Zongbo Han, Huazhu Fu, Xi Peng, Cheng Deng, Qinghua Hu, Cai Xu, Jie Wen, Di Hu, et al. Multimodal fusion on low-quality data: A comprehensive survey. arXiv preprint arXiv:2404.18947,

    2024 arXiv

  10. [30]

    Failure prediction in datacenters using unsupervised multimodal anomaly detection

    [Zhao et al., 2020] Minglu Zhao, Reo Furuhata, Mulya Agung, Hiroyuki Takizawa, and Tomoya Soma. Failure prediction in datacenters using unsupervised multimodal anomaly detection. 2020 IEEE International Conference on Big Data (Big Data), pages 3545–3549,

    2020

  11. [31]

    Improving anomaly detection in smart grid big data with knowledge graphs

    [Zhao et al., 2022] Yingying Zhao, Haoyu Sun, Naiwang Guo, Yi Wu, Yingjie Tian, and Da Zhao Cheng. Improving anomaly detection in smart grid big data with knowledge graphs. Proceedings of the 2022 6th International Confer- ence on Electronic Information Technology and Computer...

    2022

  12. [32]

    Survey on security issues of routing and anomaly detection for space information networks

    [Zhuo et al., 2021] Ming Zhuo, Leyuan Liu, Shijie Zhou, and Zhiwen Tian. Survey on security issues of routing and anomaly detection for space information networks. Scien- tific Reports, 11, 2021

    2021

  13. [2009]

    Analog and multi-modal manufacturing datasets acquired on the future factories platform

    [Harik et al., 2024] Ramy Harik, Fadi El Kalach, Jad Samaha, Devon Clark, Drew Sander, Philip Samaha, Liam Burns, Ibrahim Yousif, Victor Gadow, Theodros Tarekegne, et al. Analog and multi-modal manufacturing datasets acquired on the future factories platform. arXiv preprint ar...

    2024 arXiv

  14. [2010]

    Spatial and temporal image fusion for time series modis data and multi-sensors medium resoultion data

    [Wang et al., 2014] Xiaoqin Wang, Xiangan Chen, and Lu Li. Spatial and temporal image fusion for time series modis data and multi-sensors medium resoultion data. In 2014 IEEE Geoscience and Remote Sensing Symposium , pages 2538–2541. IEEE,

    2014

  15. [2013]

    Anomaly detection for numerical literals in knowledge graphs: A short review of approaches

    [Moghaddam et al., 2023] Farshad Bakhshandegan Moghaddam, Jens Lehmann, and Hajira Jabeen. Anomaly detection for numerical literals in knowledge graphs: A short review of approaches. 2023 IEEE Sixth In- ternational Conference on Artificial Intelligence and Knowledge Engineerin...

    2023

  16. [2014]

    Data-driven and knowledge- based predictive maintenance method for industrial robots for the production stability of intelligent manufacturing

    [Wang et al., 2023] Xiaoqiao Wang, Mingzhou Liu, Conghu Liu, Lin Ling, and Xi Zhang. Data-driven and knowledge- based predictive maintenance method for industrial robots for the production stability of intelligent manufacturing. Expert Systems with Applications, 234:121136,

    2023

  17. [2017]

    Self-supervised representation learning anomaly detection methodology based on boosting algo- rithms enhanced by data augmentation using stylegan for manufacturing imbalanced data

    [Kim et al., 2023] Yoonseok Kim, Taeheon Lee, Youngjoo Hyun, Eric Coatanea, Siren Mika, Jeonghoon Mo, and YoungJun Yoo. Self-supervised representation learning anomaly detection methodology based on boosting algo- rithms enhanced by data augmentation using stylegan for manufac...

    2023

  18. [2019]

    A survey on transfer learning

    [Pan and Yang, 2009] Sinno Jialin Pan and Qiang Yang. A survey on transfer learning. IEEE Transactions on knowl- edge and data engineering, 22(10):1345–1359,

    2009

  19. [2020]

    Multimodal fusion methods with deep neural networks and meta-information for aggression detection in surveillance

    [Jaafar and Lachiri, 2023] Noussaiba Jaafar and Zied Lachiri. Multimodal fusion methods with deep neural networks and meta-information for aggression detection in surveillance. Expert Systems with Applications , 211:118523,

    2023

  20. [2021]

    Video-based emotion recognition in the wild using deep transfer learning and score fusion

    [Kaya et al., 2017] Heysem Kaya, Furkan G¨urpınar, and Al- bert Ali Salah. Video-based emotion recognition in the wild using deep transfer learning and score fusion. Image and Vision Computing, 65:66–75,

    2017

  21. [2022]

    Towards deep industrial transfer learning for anomaly detection on time series data

    [Maschler et al., 2021] Benjamin Maschler, Tim Knodel, and Michael Weyrich. Towards deep industrial transfer learning for anomaly detection on time series data. In2021 26th IEEE International Conference on Emerging Tech- nologies and Factory Automation (ETFA) , pages 01–08. IEEE,

    2021

  22. [2023]

    Chandola and V

    [Chandola and Banerjee, 2009] V . Chandola and V . Baner- jee, A.and Kumar. Anomaly detection: A survey. ACM Computing Surveys (CSUR), 41(3):1–58,

    2009

  23. [2024]

    Time series classification using local distance- based features in multi-modal fusion networks

    [Iwana and Uchida, 2020] Brian Kenji Iwana and Seiichi Uchida. Time series classification using local distance- based features in multi-modal fusion networks. Pattern Recognition, 97:107024,

    2020

  24. [2025]

    Depression recognition based on dynamic fa- cial and vocal expression features using partial least square regression

    [Meng et al., 2013] Hongying Meng, Di Huang, Heng Wang, Hongyu Yang, Mohammed Ai-Shuraifi, and Yun- hong Wang. Depression recognition based on dynamic fa- cial and vocal expression features using partial least square regression. In Proceedings of the 3rd ACM international work...

    2013

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.