We won the Ig Nobel Prize!
Together with Nicolas Krauter, Bettina Derstroff, Christof Stönner, Efstratios Bourtsoukidis, Achim Edtbauer, Jochen Wulf, Thomas Klüpfel, Stefan Kramer, and Jonathan […]
I am senior lecturer at the School of Computer Science of the University of Auckland, CTO of enviPath, and lead the Machine Learning Group at UoA. My main research area is machine learning and its application to bioinformatics, cheminformatics, and computational sustainability. Before joining the University of Auckland in 2017, I did a PostDoc at University of Mainz, Germany, and a PhD at Technical University of Munich, Germany. My current research focuses on diverse areas ranging from reliability of machine learning algorithms, particularly in respect to adversarial learning and bias to applications in epidemology, SLAM, environmental research, and cybersecurity.
I am currently looking for PhD, Honours, or Masters students, if you are interested in any of my research areas, contact me by mail. Also check the Supervision page for funding information.
Journal Articles
Bensemann, Joshua; Cheena, Hasnain; Huang, David Tse Juang; Broadbendt, Elizabeth; Williams, Jonathan; Wicker, Jörg
From What You See to What We Smell: Linking Human Emotions to Biomarkers in Breath Journal Article Forthcoming
In: Forthcoming.
@article{bensemann2022from,
title = {From What You See to What We Smell: Linking Human Emotions to Biomarkers in Breath},
author = {Joshua Bensemann and Hasnain Cheena and David Tse Juang Huang and Elizabeth Broadbendt and Jonathan Williams and J\"{o}rg Wicker},
year = {2022},
date = {2022-12-01},
abstract = {Breath collection is a non-invasive method for monitoring biological processes occurring within the human body. Prior studies have extended these methods to observe the general processes occurring in groups of humans and are able to link them to what those groups are collectively experiencing. However, previous work lacked an objective measure of emotional stimuli. In this research, we applied machine learning techniques to breath data collected from cinema audiences to find associations between the biomarkers in the crowd's breath and both the audio-visual stimuli and thematic events of the movie.
This analysis enabled us to make direct links between what the group was experiencing and their biological response to that experience. To achieve this, we first extracted visual and auditory features from a movie and compared it to the biomarkers in the crowd's breath, using both regression and pattern mining techniques. Our results supported the theory that a crowd's collective experience has a direct correlation to the biomarkers in the crowd's breath. Consequently, these findings suggest that visual and auditory experiences have predictable effects on the human body that can be monitored without the requirement of expensive and/or invasive neuroimaging techniques.},
keywords = {},
pubstate = {forthcoming},
tppubtype = {article}
}
Breath collection is a non-invasive method for monitoring biological processes occurring within the human body. Prior studies have extended these methods to observe the general processes occurring in groups of humans and are able to link them to what those groups are collectively experiencing. However, previous work lacked an objective measure of emotional stimuli. In this research, we applied machine learning techniques to breath data collected from cinema audiences to find associations between the biomarkers in the crowd's breath and both the audio-visual stimuli and thematic events of the movie.
This analysis enabled us to make direct links between what the group was experiencing and their biological response to that experience. To achieve this, we first extracted visual and auditory features from a movie and compared it to the biomarkers in the crowd's breath, using both regression and pattern mining techniques. Our results supported the theory that a crowd's collective experience has a direct correlation to the biomarkers in the crowd's breath. Consequently, these findings suggest that visual and auditory experiences have predictable effects on the human body that can be monitored without the requirement of expensive and/or invasive neuroimaging techniques.
This analysis enabled us to make direct links between what the group was experiencing and their biological response to that experience. To achieve this, we first extracted visual and auditory features from a movie and compared it to the biomarkers in the crowd's breath, using both regression and pattern mining techniques. Our results supported the theory that a crowd's collective experience has a direct correlation to the biomarkers in the crowd's breath. Consequently, these findings suggest that visual and auditory experiences have predictable effects on the human body that can be monitored without the requirement of expensive and/or invasive neuroimaging techniques.
Pullar-Strecker, Zac; Dost, Katharina; Frank, Eibe; Wicker, Jörg
Hitting the Target: Stopping Active Learning at the Cost-Based Optimum Journal Article
In: arXiv preprint, 2021.
@article{pullar-strecker2021hitting,
title = {Hitting the Target: Stopping Active Learning at the Cost-Based Optimum},
author = {Zac Pullar-Strecker and Katharina Dost and Eibe Frank and J\"{o}rg Wicker},
url = {https://arxiv.org/abs/2110.03802},
year = {2021},
date = {2021-10-08},
urldate = {2021-10-08},
journal = {arXiv preprint},
abstract = {Active learning allows machine learning models to be trained using fewer labels while retaining similar performance to traditional fully supervised learning. An active learner selects the most informative data points, requests their labels, and retrains itself. While this approach is promising, it leaves an open problem of how to determine when the model is `good enough' without the additional labels required for traditional evaluation. In the past, different stopping criteria have been proposed aiming to identify the optimal stopping point. However, optimality can only be expressed as a domain-dependent trade-off between accuracy and the number of labels, and no criterion is superior in all applications. This paper is the first to give actionable advice to practitioners on what stopping criteria they should use in a given real-world scenario. We contribute the first large-scale comparison of stopping criteria, using a cost measure to quantify the accuracy/label trade-off, public implementations of all stopping criteria we evaluate, and an open-source framework for evaluating stopping criteria. Our research enables practitioners to substantially reduce labelling costs by utilizing the stopping criterion which best suits their domain.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Active learning allows machine learning models to be trained using fewer labels while retaining similar performance to traditional fully supervised learning. An active learner selects the most informative data points, requests their labels, and retrains itself. While this approach is promising, it leaves an open problem of how to determine when the model is `good enough' without the additional labels required for traditional evaluation. In the past, different stopping criteria have been proposed aiming to identify the optimal stopping point. However, optimality can only be expressed as a domain-dependent trade-off between accuracy and the number of labels, and no criterion is superior in all applications. This paper is the first to give actionable advice to practitioners on what stopping criteria they should use in a given real-world scenario. We contribute the first large-scale comparison of stopping criteria, using a cost measure to quantify the accuracy/label trade-off, public implementations of all stopping criteria we evaluate, and an open-source framework for evaluating stopping criteria. Our research enables practitioners to substantially reduce labelling costs by utilizing the stopping criterion which best suits their domain.
Tam, Jason; Lorsbach, Tim; Schmidt, Sebastian; Wicker, Jörg
Holistic Evaluation of Biodegradation Pathway Prediction: Assessing Multi-Step Reactions and Intermediate Products Journal Article
In: Journal of Cheminformatics, vol. 13, no. 1, pp. 63, 2021.
Abstract | Links | BibTeX | Altmetric
@article{tam2021holisticb,
title = {Holistic Evaluation of Biodegradation Pathway Prediction: Assessing Multi-Step Reactions and Intermediate Products},
author = {Jason Tam and Tim Lorsbach and Sebastian Schmidt and J\"{o}rg Wicker},
url = {https://jcheminf.biomedcentral.com/articles/10.1186/s13321-021-00543-x
https://chemrxiv.org/articles/preprint/Holistic_Evaluation_of_Biodegradation_Pathway_Prediction_Assessing_Multi-Step_Reactions_and_Intermediate_Products/14315963
https://dx.doi.org/10.26434/chemrxiv.14315963},
doi = {10.1186/s13321-021-00543-x},
year = {2021},
date = {2021-09-03},
urldate = {2021-09-03},
journal = {Journal of Cheminformatics},
volume = {13},
number = {1},
pages = {63},
abstract = {The prediction of metabolism and biotransformation pathways of xenobiotics is a highly desired tool in environmental sciences, drug discovery, and (eco)toxicology. Several systems predict single transformation steps or complete pathways as series of parallel and subsequent steps. Their performance is commonly evaluated on the level of a single transformation step. Such an approach cannot account for some specific challenges that are caused by specific properties of biotransformation experiments. That is, missing transformation products in the reference data that occur only in low concentrations, e.g. transient intermediates or higher-generation metabolites. Furthermore, some rule-based prediction systems evaluate the performance only based on the defined set of transformation rules. Therefore, the performance of these models cannot be directly compared. In this paper, we introduce a new evaluation framework that extends the evaluation of biotransformation prediction from single transformations to whole pathways, taking into account multiple generations of metabolites. We introduce a procedure to address transient intermediates and propose a weighted scoring system that acknowledges the uncertainty of higher-generation metabolites. We implemented this framework in enviPath and demonstrate its strict performance metrics on predictions of in vitro biotransformation and degradation of xenobiotics in soil. Our approach is model-agnostic and can be transferred to other prediction systems. It is also capable of revealing knowledge gaps in terms of incompletely defined sets of transformation rules.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The prediction of metabolism and biotransformation pathways of xenobiotics is a highly desired tool in environmental sciences, drug discovery, and (eco)toxicology. Several systems predict single transformation steps or complete pathways as series of parallel and subsequent steps. Their performance is commonly evaluated on the level of a single transformation step. Such an approach cannot account for some specific challenges that are caused by specific properties of biotransformation experiments. That is, missing transformation products in the reference data that occur only in low concentrations, e.g. transient intermediates or higher-generation metabolites. Furthermore, some rule-based prediction systems evaluate the performance only based on the defined set of transformation rules. Therefore, the performance of these models cannot be directly compared. In this paper, we introduce a new evaluation framework that extends the evaluation of biotransformation prediction from single transformations to whole pathways, taking into account multiple generations of metabolites. We introduce a procedure to address transient intermediates and propose a weighted scoring system that acknowledges the uncertainty of higher-generation metabolites. We implemented this framework in enviPath and demonstrate its strict performance metrics on predictions of in vitro biotransformation and degradation of xenobiotics in soil. Our approach is model-agnostic and can be transferred to other prediction systems. It is also capable of revealing knowledge gaps in terms of incompletely defined sets of transformation rules.
Chang, Luke; Dost, Katharina; Zhai, Kaiqi; Demontis, Ambra; Roli, Fabio; Dobbie, Gillian; Wicker, Jörg
Intriguing Usage of Applicability Domain: Lessons from Cheminformatics Applied to Adversarial Learning Journal Article
In: arxiv, vol. 2105.00495, 2021, (preprint).
@article{chang2021intriguing,
title = {Intriguing Usage of Applicability Domain: Lessons from Cheminformatics Applied to Adversarial Learning},
author = {Luke Chang and Katharina Dost and Kaiqi Zhai and Ambra Demontis and Fabio Roli and Gillian Dobbie and J\"{o}rg Wicker},
url = {https://arxiv.org/abs/2105.00495},
year = {2021},
date = {2021-05-02},
journal = {arxiv},
volume = {2105.00495},
abstract = {Defending machine learning models from adversarial attacks is still a challenge: none of the robust models is utterly immune to adversarial examples to date. Different defences have been proposed; however, most of them are tailored to particular ML models and adversarial attacks, therefore their effectiveness and applicability are strongly limited. A similar problem plagues cheminformatics: Quantitative Structure-Activity Relationship (QSAR) models struggle to predict biological activity for the entire chemical space because they are trained on a very limited amount of compounds with known effects. This problem is relieved with a technique called Applicability Domain (AD), which rejects the unsuitable compounds for the model. Adversarial examples are intentionally crafted inputs that exploit the blind spots which the model has not learned to classify, and adversarial defences try to make the classifier more robust by covering these blind spots. There is an apparent similarity between AD and adversarial defences. Inspired by the concept of AD, we propose a multi-stage data-driven defence that is testing for: Applicability: abnormal values, namely inputs not compliant with the intended use case of the model; Reliability: samples far from the training data; and Decidability: samples whose predictions contradict the predictions of their neighbours. It can be applied to any classification model and is not limited to specific types of adversarial attacks. With an empirical analysis, this paper demonstrates how Applicability Domain can effectively reduce the vulnerability of ML models to adversarial examples. },
note = {preprint},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Defending machine learning models from adversarial attacks is still a challenge: none of the robust models is utterly immune to adversarial examples to date. Different defences have been proposed; however, most of them are tailored to particular ML models and adversarial attacks, therefore their effectiveness and applicability are strongly limited. A similar problem plagues cheminformatics: Quantitative Structure-Activity Relationship (QSAR) models struggle to predict biological activity for the entire chemical space because they are trained on a very limited amount of compounds with known effects. This problem is relieved with a technique called Applicability Domain (AD), which rejects the unsuitable compounds for the model. Adversarial examples are intentionally crafted inputs that exploit the blind spots which the model has not learned to classify, and adversarial defences try to make the classifier more robust by covering these blind spots. There is an apparent similarity between AD and adversarial defences. Inspired by the concept of AD, we propose a multi-stage data-driven defence that is testing for: Applicability: abnormal values, namely inputs not compliant with the intended use case of the model; Reliability: samples far from the training data; and Decidability: samples whose predictions contradict the predictions of their neighbours. It can be applied to any classification model and is not limited to specific types of adversarial attacks. With an empirical analysis, this paper demonstrates how Applicability Domain can effectively reduce the vulnerability of ML models to adversarial examples.
Stepišnik, Tomaž; Škrlj, Blaž; Wicker, Jörg; Kocev, Dragi
A comprehensive comparison of molecular feature representations for use in predictive modeling Journal Article
In: Computers in Biology and Medicine, vol. 130, pp. 104197, 2021, ISSN: 0010-4825.
Abstract | Links | BibTeX | Altmetric
@article{stepisnik2021comprehensive,
title = {A comprehensive comparison of molecular feature representations for use in predictive modeling},
author = {Toma\v{z} Stepi\v{s}nik and Bla\v{z} \v{S}krlj and J\"{o}rg Wicker and Dragi Kocev},
url = {http://www.sciencedirect.com/science/article/pii/S001048252030528X},
doi = {10.1016/j.compbiomed.2020.104197},
issn = {0010-4825},
year = {2021},
date = {2021-03-01},
journal = {Computers in Biology and Medicine},
volume = {130},
pages = {104197},
abstract = {Machine learning methods are commonly used for predicting molecular properties to accelerate material and drug design. An important part of this process is deciding how to represent the molecules. Typically, machine learning methods expect examples represented by vectors of values, and many methods for calculating molecular feature representations have been proposed. In this paper, we perform a comprehensive comparison of different molecular features, including traditional methods such as fingerprints and molecular descriptors, and recently proposed learnable representations based on neural networks. Feature representations are evaluated on 11 benchmark datasets, used for predicting properties and measures such as mutagenicity, melting points, activity, solubility, and IC50. Our experiments show that several molecular features work similarly well over all benchmark datasets. The ones that stand out most are Spectrophores, which give significantly worse performance than other features on most datasets. Molecular descriptors from the PaDEL library seem very well suited for predicting physical properties of molecules. Despite their simplicity, MACCS fingerprints performed very well overall. The results show that learnable representations achieve competitive performance compared to expert based representations. However, task-specific representations (graph convolutions and Weave methods) rarely offer any benefits, even though they are computationally more demanding. Lastly, combining different molecular feature representations typically does not give a noticeable improvement in performance compared to individual feature representations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Machine learning methods are commonly used for predicting molecular properties to accelerate material and drug design. An important part of this process is deciding how to represent the molecules. Typically, machine learning methods expect examples represented by vectors of values, and many methods for calculating molecular feature representations have been proposed. In this paper, we perform a comprehensive comparison of different molecular features, including traditional methods such as fingerprints and molecular descriptors, and recently proposed learnable representations based on neural networks. Feature representations are evaluated on 11 benchmark datasets, used for predicting properties and measures such as mutagenicity, melting points, activity, solubility, and IC50. Our experiments show that several molecular features work similarly well over all benchmark datasets. The ones that stand out most are Spectrophores, which give significantly worse performance than other features on most datasets. Molecular descriptors from the PaDEL library seem very well suited for predicting physical properties of molecules. Despite their simplicity, MACCS fingerprints performed very well overall. The results show that learnable representations achieve competitive performance compared to expert based representations. However, task-specific representations (graph convolutions and Weave methods) rarely offer any benefits, even though they are computationally more demanding. Lastly, combining different molecular feature representations typically does not give a noticeable improvement in performance compared to individual feature representations.
Inproceedings
Dost, Katharina; Duncanson, Hamish; Ziogas, Ioannis; Riddle, Pat; Wicker, Jörg
Divide and Imitate: Multi-Cluster Identification and Mitigation of Selection Bias Inproceedings
In: 26th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2022), pp. 149–160, Springer-Verlag, Berlin, Heidelberg, 2022, ISBN: 978-3-031-05935-3.
Abstract | Links | BibTeX | Altmetric
@inproceedings{dost2022divide,
title = {Divide and Imitate: Multi-Cluster Identification and Mitigation of Selection Bias},
author = {Katharina Dost and Hamish Duncanson and Ioannis Ziogas and Pat Riddle and J\"{o}rg Wicker},
doi = {10.1007/978-3-031-05936-0_12},
isbn = {978-3-031-05935-3},
year = {2022},
date = {2022-05-16},
urldate = {2022-05-16},
booktitle = {26th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2022)},
pages = {149\textendash160},
publisher = {Springer-Verlag},
address = {Berlin, Heidelberg},
abstract = {Machine Learning can help overcome human biases in decision making by focusing on purely logical conclusions based on the training data. If the training data is biased, however, that bias will be transferred to the model and remains undetected as the performance is validated on a test set drawn from the same biased distribution. Existing strategies for selection bias identification and mitigation generally rely on some sort of knowledge of the bias or the ground-truth. An exception is the Imitate algorithm that assumes no knowledge but comes with a strong limitation: It can only model datasets with one normally distributed cluster per class. In this paper, we introduce a novel algorithm, Mimic, which uses Imitate as a building block but relaxes this limitation. By allowing mixtures of multivariate Gaussians, our technique is able to model multi-cluster datasets and provide solutions for a substantially wider set of problems. Experiments confirm that Mimic not only identifies potential biases in multi-cluster datasets which can be corrected early on but also improves classifier performance.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Machine Learning can help overcome human biases in decision making by focusing on purely logical conclusions based on the training data. If the training data is biased, however, that bias will be transferred to the model and remains undetected as the performance is validated on a test set drawn from the same biased distribution. Existing strategies for selection bias identification and mitigation generally rely on some sort of knowledge of the bias or the ground-truth. An exception is the Imitate algorithm that assumes no knowledge but comes with a strong limitation: It can only model datasets with one normally distributed cluster per class. In this paper, we introduce a novel algorithm, Mimic, which uses Imitate as a building block but relaxes this limitation. By allowing mixtures of multivariate Gaussians, our technique is able to model multi-cluster datasets and provide solutions for a substantially wider set of problems. Experiments confirm that Mimic not only identifies potential biases in multi-cluster datasets which can be corrected early on but also improves classifier performance.
Graffeuille, Olivier; Koh, Yun Sing; Wicker, Jörg; Lehmann, Moritz
Semi-Supervised Conditional Density Estimation with Wasserstein Laplacian Regularisation Inproceedings Forthcoming
In: Proceeding of the Thirty-Sixth AAAI Conference on Artificial Intelligence, Forthcoming.
@inproceedings{graffeuille2022semi,
title = {Semi-Supervised Conditional Density Estimation with Wasserstein Laplacian Regularisation},
author = {Olivier Graffeuille and Yun Sing Koh and J\"{o}rg Wicker and Moritz Lehmann},
year = {2022},
date = {2022-02-22},
booktitle = {Proceeding of the Thirty-Sixth AAAI Conference on Artificial Intelligence},
abstract = {Conditional Density Estimation (CDE) has wide-reaching applicability to various real-world problems, such as spatial density estimation and environmental modelling. CDE estimates the probability density of a random variable rather than a single value and can thus model uncertainty and inverse problems. This task is inherently more complex than regression, and many algorithms suffer from overfitting, particularly when modelled with few labelled data points. For applications where unlabelled data is abundant but labelled data is scarce, we propose Wasserstein Laplacian Regularisation, a semi-supervised learning framework that allows CDE algorithms to leverage these unlabelled data. The framework minimises an objective function which ensures that the learned model is smooth along the manifold of the underlying data, as measured by Wasserstein distance. When applying our framework to Mixture Density Networks, the resulting semi-supervised algorithm can achieve similar performance to a supervised model with up to three times as many labelled data points on baseline datasets. We additionally apply our technique to the problem of remote sensing for chlorophyll-a estimation in inland waters.},
keywords = {},
pubstate = {forthcoming},
tppubtype = {inproceedings}
}
Conditional Density Estimation (CDE) has wide-reaching applicability to various real-world problems, such as spatial density estimation and environmental modelling. CDE estimates the probability density of a random variable rather than a single value and can thus model uncertainty and inverse problems. This task is inherently more complex than regression, and many algorithms suffer from overfitting, particularly when modelled with few labelled data points. For applications where unlabelled data is abundant but labelled data is scarce, we propose Wasserstein Laplacian Regularisation, a semi-supervised learning framework that allows CDE algorithms to leverage these unlabelled data. The framework minimises an objective function which ensures that the learned model is smooth along the manifold of the underlying data, as measured by Wasserstein distance. When applying our framework to Mixture Density Networks, the resulting semi-supervised algorithm can achieve similar performance to a supervised model with up to three times as many labelled data points on baseline datasets. We additionally apply our technique to the problem of remote sensing for chlorophyll-a estimation in inland waters.
Poonawala-Lohani, Nooriyan; Riddle, Pat; Adnan, Mehnaz; Wicker, Jörg
A Novel Approach for Time Series Forecasting of Influenza-like Illness Using a Regression Chain Method Inproceedings
In: Altman, Russ; Dunker, Keith; Hunter, Lawrence; Ritchie, Marylyn; Murray, Tiffany; Klein, Teri (Ed.): Pacific Symposium on Biocomputing, pp. 301-312, 2022.
Abstract | Links | BibTeX | Altmetric
@inproceedings{poonawala-lohani2022novel,
title = {A Novel Approach for Time Series Forecasting of Influenza-like Illness Using a Regression Chain Method},
author = {Nooriyan Poonawala-Lohani and Pat Riddle and Mehnaz Adnan and J\"{o}rg Wicker},
editor = {Russ Altman and Keith Dunker and Lawrence Hunter and Marylyn Ritchie and Tiffany Murray and Teri Klein},
url = {https://www.worldscientific.com/doi/abs/10.1142/9789811250477_0028
http://psb.stanford.edu/psb-online/proceedings/psb22/poorawala-lohani.pdf},
doi = {10.1142/9789811250477_0028},
year = {2022},
date = {2022-01-03},
urldate = {2022-01-03},
booktitle = {Pacific Symposium on Biocomputing},
volume = {27},
pages = {301-312},
abstract = {Influenza is a communicable respiratory illness that can cause serious public health hazards. Due to its huge threat to the community, accurate forecasting of Influenza-like-illness (ILI) can diminish the impact of an influenza season by enabling early public health interventions. Current forecasting models are limited in their performance, particularly when using a longer forecasting window. To support better forecasts over a longer forecasting window, we propose to use additional features such as weather data. Commonly used methods to fore-cast ILI, including statistical methods such as ARIMA, limit prediction performance when using additional data sources that might have complex non-linear associations with ILI incidence. This paper proposes a novel time series forecasting method, Randomized Ensembles of Auto-regression chains (Reach). Reach implements an ensemble of random chains for multi-step time series forecasting. This new approach is evaluated on ILI case counts in Auckland, New Zealand from the years 2015-2018 and compared to other standard methods. The results demonstrate that the proposed method performed better than baseline methods when applied to this multi-variate time series forecasting problem.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Influenza is a communicable respiratory illness that can cause serious public health hazards. Due to its huge threat to the community, accurate forecasting of Influenza-like-illness (ILI) can diminish the impact of an influenza season by enabling early public health interventions. Current forecasting models are limited in their performance, particularly when using a longer forecasting window. To support better forecasts over a longer forecasting window, we propose to use additional features such as weather data. Commonly used methods to fore-cast ILI, including statistical methods such as ARIMA, limit prediction performance when using additional data sources that might have complex non-linear associations with ILI incidence. This paper proposes a novel time series forecasting method, Randomized Ensembles of Auto-regression chains (Reach). Reach implements an ensemble of random chains for multi-step time series forecasting. This new approach is evaluated on ILI case counts in Auckland, New Zealand from the years 2015-2018 and compared to other standard methods. The results demonstrate that the proposed method performed better than baseline methods when applied to this multi-variate time series forecasting problem.
Kim, Jonathan; Urschler, Martin; Riddle, Pat; Wicker, Jörg
SymbioLCD: Ensemble-Based Loop Closure Detection using CNN-Extracted Objects and Visual Bag-of-Words Inproceedings
In: 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 5425-5425, 2021.
Abstract | Links | BibTeX | Altmetric
@inproceedings{kim2021symbiolcd,
title = {SymbioLCD: Ensemble-Based Loop Closure Detection using CNN-Extracted Objects and Visual Bag-of-Words},
author = {Jonathan Kim and Martin Urschler and Pat Riddle and J\"{o}rg Wicker},
url = {https://ieeexplore.ieee.org/abstract/document/9636622
http://arxiv.org/abs/2110.11491},
doi = {10.1109/IROS51168.2021.9636622},
year = {2021},
date = {2021-09-27},
urldate = {2021-09-27},
booktitle = {2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages = {5425-5425},
abstract = {Loop closure detection is an essential tool of Simultaneous Localization and Mapping (SLAM) to minimize drift in its localization. Many state-of-the-art loop closure detection (LCD) algorithms use visual Bag-of-Words (vBoW), which is robust against partial occlusions in a scene but cannot perceive the semantics or spatial relationships between feature points. CNN object extraction can address those issues, by providing semantic labels and spatial relationships between objects in a scene. Previous work has mainly focused on replacing vBoW with CNN derived features.
In this paper we propose SymbioLCD, a novel ensemble-based LCD that utilizes both CNN-extracted objects and vBoW features for LCD candidate prediction. When used in tandem, the added elements of object semantics and spatial-awareness creates a more robust and symbiotic loop closure detection system. The proposed SymbioLCD uses scale-invariant spatial and semantic matching, Hausdorff distance with temporal constraints, and a Random Forest that utilizes combined information from both CNN-extracted objects and vBoW features for predicting accurate loop closure candidates. Evaluation of the proposed method shows it outperforms other Machine Learning (ML) algorithms - such as SVM, Decision Tree and Neural Network, and demonstrates that there is a strong symbiosis between CNN-extracted object information and vBoW features which assists accurate LCD candidate prediction. Furthermore, it is able to perceive loop closure candidates earlier than state-of-the-art SLAM algorithms, utilizing added spatial and semantic information from CNN-extracted objects.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Loop closure detection is an essential tool of Simultaneous Localization and Mapping (SLAM) to minimize drift in its localization. Many state-of-the-art loop closure detection (LCD) algorithms use visual Bag-of-Words (vBoW), which is robust against partial occlusions in a scene but cannot perceive the semantics or spatial relationships between feature points. CNN object extraction can address those issues, by providing semantic labels and spatial relationships between objects in a scene. Previous work has mainly focused on replacing vBoW with CNN derived features.
In this paper we propose SymbioLCD, a novel ensemble-based LCD that utilizes both CNN-extracted objects and vBoW features for LCD candidate prediction. When used in tandem, the added elements of object semantics and spatial-awareness creates a more robust and symbiotic loop closure detection system. The proposed SymbioLCD uses scale-invariant spatial and semantic matching, Hausdorff distance with temporal constraints, and a Random Forest that utilizes combined information from both CNN-extracted objects and vBoW features for predicting accurate loop closure candidates. Evaluation of the proposed method shows it outperforms other Machine Learning (ML) algorithms - such as SVM, Decision Tree and Neural Network, and demonstrates that there is a strong symbiosis between CNN-extracted object information and vBoW features which assists accurate LCD candidate prediction. Furthermore, it is able to perceive loop closure candidates earlier than state-of-the-art SLAM algorithms, utilizing added spatial and semantic information from CNN-extracted objects.
In this paper we propose SymbioLCD, a novel ensemble-based LCD that utilizes both CNN-extracted objects and vBoW features for LCD candidate prediction. When used in tandem, the added elements of object semantics and spatial-awareness creates a more robust and symbiotic loop closure detection system. The proposed SymbioLCD uses scale-invariant spatial and semantic matching, Hausdorff distance with temporal constraints, and a Random Forest that utilizes combined information from both CNN-extracted objects and vBoW features for predicting accurate loop closure candidates. Evaluation of the proposed method shows it outperforms other Machine Learning (ML) algorithms - such as SVM, Decision Tree and Neural Network, and demonstrates that there is a strong symbiosis between CNN-extracted object information and vBoW features which assists accurate LCD candidate prediction. Furthermore, it is able to perceive loop closure candidates earlier than state-of-the-art SLAM algorithms, utilizing added spatial and semantic information from CNN-extracted objects.
