No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Tackle Cognitive Biases in Videosurveillance
ResearchGate has not been able to resolve any citations for this publication.
Advances in machine learning (ML) and artificial intelligence (AI) present an opportunity to build better tools and solutions to help address some of the world’s most pressing challenges, and deliver positive social impact in accordance with the priorities outlined in the United Nations’ 17 Sustainable Development Goals (SDGs). The AI for Social Good (AI4SG) movement aims to establish interdisciplinary partnerships centred around AI applications towards SDGs. We provide a set of guidelines for establishing successful long-term collaborations between AI researchers and application-domain experts, relate them to existing AI4SG projects and identify key opportunities for future AI applications targeted towards social good. The AI for Social Good movement aims to apply AI/ML tools to help in delivering on the United Nations’ sustainable development goals (SDGs). Here, the authors identify the challenges and propose guidelines for designing and implementing successful partnerships between AI researchers and application - domain experts.
This work explores the biases in learning processes based on deep neural network architectures through a case study in gender detection from face images. We employ two gender detection models based on popular deep neural networks. We present a comprehensive analysis of bias effects when using an unbalanced training dataset on the features learned by the models. We show how ethnic attributes impact in the activations of gender detection models based on face images. We finally propose InsideBias, a novel method to detect biased models. InsideBias is based on how the models represent the information instead of how they perform, which is the normal practice in other existing methods for bias detection. Our strategy with InsideBias allows to detect biased models with very few samples (only 15 images in our case study). Our experiments include 72K face images from 24K identities and 3 ethnic groups.
Social data in digital form—including user-generated content, expressed or implicit relations between people, and behavioral traces—are at the core of popular applications and platforms, driving the research agenda of many researchers. The promises of social data are many, including understanding “what the world thinks” about a social issue, brand, celebrity, or other entity, as well as enabling better decision-making in a variety of fields including public policy, healthcare, and economics. Many academics and practitioners have warned against the naïve usage of social data. There are biases and inaccuracies occurring at the source of the data, but also introduced during processing. There are methodological limitations and pitfalls, as well as ethical boundaries and unexpected consequences that are often overlooked. This paper recognizes the rigor with which these issues are addressed by different researchers varies across a wide range. We identify a variety of menaces in the practices around social data use, and organize them in a framework that helps to identify them. “For your own sanity, you have to remember that not all problems can be solved. Not all problems can be solved, but all problems can be illuminated.” –Ursula Franklin1 “For your own sanity, you have to remember that not all problems can be solved. Not all problems can be solved, but all problems can be illuminated.” –Ursula Franklin1
Recent awareness of the impacts of bias in AI algorithms raises the risk for companies to deploy such algorithms, especially because the algorithms may not be explainable in the same way that non-AI algorithms are. Even with careful review of the algorithms and data sets, it may not be possible to delete all unwanted bias, particularly because AI systems learn from historical data, which encodes historical biases. In this paper, we propose a set of processes that companies can use to mitigate and manage three general classes of bias: those related to mapping the business intent into the AI implementation, those that arise due to the distribution of samples used for training, and those that are present in individual input samples. While there may be no simple or complete solution to this issue, best practices can be used to reduce the effects of bias on algorithmic outcomes.
Combining big data and machine learning algorithms, the power of automatic decision tools induces as much hope as fear. Many recently enacted European legislation (GDPR) and French laws attempt to regulate the use of these tools. Leaving aside the well-identified problems of data confidentiality and impediments to competition, we focus on the risks of discrimination, the problems of transparency and the quality of algorithmic decisions. The detailed perspective of the legal texts, faced with the complexity and opacity of the learning algorithms, reveals the need for important technological disruptions for the detection or reduction of the discrimination risk, and for addressing the right to obtain an explanation of the automatic decision. Since trust of the developers and above all of the users (citizens, litigants, customers) is essential, algorithms exploiting personal data must be deployed in a strict ethical framework. In conclusion, to answer this need, we list some ways of controls to be developed: institutional control, ethical charter, external audit attached to the issue of a label.
Biases in AI and machine learning algorithms are presented and analyzed through two issues management frameworks with the aim of showing how ethical problems and dilemmas can evolve. While “the singularity” concept in AI is presently more predictive than actual, both benefits and damage that can result by failure to consider biases in the design and development of AI. Inclusivity and stakeholder awareness regarding potential ethical risks and issues need to be identified during the design of AI algorithms to ensure that the most vulnerable in societies are protected from harm.
We consider the paradigm of a black box AI system that makes life-critical decisions. We propose an "arguing machines" framework that pairs the primary AI system with a secondary one that is independently trained to perform the same task. We show that disagreement between the two systems, without any knowledge of underlying system design or operation, is sufficient to arbitrarily improve the accuracy of the overall decision pipeline given human supervision over disagreements. We demonstrate this system in two applications: (1) an illustrative example of image classification and (2) on large-scale real-world semi-autonomous driving data. For the first application, we apply this framework to image classification achieving a reduction from 8.0% to 2.8% top-5 error on ImageNet. For the second application, we apply this framework to Tesla Autopilot and demonstrate the ability to predict 90.4% of system disengagements that were labeled by human annotators as challenging and needing human supervision.
A novel framework for anomaly detection in crowded scenes is presented. Three properties are identified as important for the design of a localized video representation suitable for anomaly detection in such scenes: (1) joint modeling of appearance and dynamics of the scene, and the abilities to detect (2) temporal, and (3) spatial abnormalities. The model for normal crowd behavior is based on mixtures of dynamic textures and outliers under this model are labeled as anomalies. Temporal anomalies are equated to events of low-probability, while spatial anomalies are handled using discriminant saliency. An experimental evaluation is conducted with a new dataset of crowded scenes, composed of 100 video sequences and five well defined abnormality categories. The proposed representation is shown to outperform various state of the art anomaly detection techniques.
An approach to the construction of classifiers from imbalanced datasets is
described. A dataset is imbalanced if the classification categories are not
approximately equally represented. Often real-world data sets are predominately
composed of "normal" examples with only a small percentage of "abnormal" or
"interesting" examples. It is also the case that the cost of misclassifying an
abnormal (interesting) example as a normal example is often much higher than
the cost of the reverse error. Under-sampling of the majority (normal) class
has been proposed as a good means of increasing the sensitivity of a classifier
to the minority class. This paper shows that a combination of our method of
over-sampling the minority (abnormal) class and under-sampling the majority
(normal) class can achieve better classifier performance (in ROC space) than
only under-sampling the majority class. This paper also shows that a
combination of our method of over-sampling the minority class and
under-sampling the majority class can achieve better classifier performance (in
ROC space) than varying the loss ratios in Ripper or class priors in Naive
Bayes. Our method of over-sampling the minority class involves creating
synthetic minority class examples. Experiments are performed using C4.5, Ripper
and a Naive Bayes classifier. The method is evaluated using the area under the
Receiver Operating Characteristic curve (AUC) and the ROC convex hull strategy.
With the widespread use of artificial intelligence (AI) systems and applications in our everyday lives, accounting for fairness has gained significant importance in designing and engineering of such systems. AI systems can be used in many sensitive environments to make important and life-changing decisions; thus, it is crucial to ensure that these decisions do not reflect discriminatory behavior toward certain groups or populations. More recently some work has been developed in traditional machine learning and deep learning that address such challenges in different subdomains. With the commercialization of these systems, researchers are becoming more aware of the biases that these applications can contain and are attempting to address them. In this survey, we investigated different real-world applications that have shown biases in various ways, and we listed different sources of biases that can affect AI applications. We then created a taxonomy for fairness definitions that machine learning researchers have defined to avoid the existing bias in AI systems. In addition to that, we examined different domains and subdomains in AI showing what researchers have observed with regard to unfair outcomes in the state-of-the-art methods and ways they have tried to address them. There are still many future directions and solutions that can be taken to mitigate the problem of bias in AI systems. We are hoping that this survey will motivate researchers to tackle these issues in the near future by observing existing work in their respective fields.
Algorithm fairness has started to attract the attention of researchers in AI, Software Engineering and Law communities, with more than twenty different notions of fairness proposed in the last few years. Yet, there is no clear agreement on which definition to apply in each situation. Moreover, the detailed differences between multiple definitions are difficult to grasp. To address this issue, this paper collects the most prominent definitions of fairness for the algorithmic classification problem, explains the rationale behind these definitions, and demonstrates each of them on a single unifying case-study. Our analysis intuitively explains why the same case can be considered fair according to some definitions and unfair according to others.
Despite widespread adoption, machine learning models remain mostly black boxes. Understanding the reasons behind predictions is, however, quite important in assessing trust in a model. Trust is fundamental if one plans to take action based on a prediction, or when choosing whether or not to deploy a new model. Such understanding further provides insights into the model, which can be used to turn an untrustworthy model or prediction into a trustworthy one. In this work, we propose LIME, a novel explanation technique that explains the predictions of any classifier in an interpretable and faithful manner, by learning an interpretable model locally around the prediction. We further propose a method to explain models by presenting representative individual predictions and their explanations in a non-redundant way, framing the task as a submodular optimization problem. We demonstrate the flexibility of these methods by explaining different models for text (e.g. random forests) and image classification (e.g. neural networks). The usefulness of explanations is shown via novel experiments, both simulated and with human subjects. Our explanations empower users in various scenarios that require trust: deciding if one should trust a prediction, choosing between models, improving an untrustworthy classifier, and detecting why a classifier should not be trusted.
Despite widespread adoption, machine learning models remain mostly black boxes. Understanding the reasons behind predictions is, however, quite important in assessing trust, which is fundamental if one plans to take action based on a prediction, or when choosing whether to deploy a new model. Such understanding also provides insights into the model, which can be used to transform an untrustworthy model or prediction into a trustworthy one.
In this work, we propose LIME, a novel explanation technique that explains the predictions of any classifier in an interpretable and faithful manner, by learning an interpretable model locally varound the prediction. We also propose a method to explain models by presenting representative individual predictions and their explanations in a non-redundant way, framing the task as a submodular optimization problem. We demonstrate the flexibility of these methods by explaining different models for text (e.g. random forests) and image classification (e.g. neural networks). We show the utility of explanations via novel experiments, both simulated and with human subjects, on various scenarios that require trust: deciding if one should trust a prediction, choosing between models, improving an untrustworthy classifier, and identifying why a classifier should not be trusted.
Despite widespread adoption, machine learning models remain mostly black boxes. Understanding the reasons behind predictions is, however, quite important in assessing trust in a model. Trust is fundamental if one plans to take action based on a prediction, or when choosing whether or not to deploy a new model. Such understanding further provides insights into the model, which can be used to turn an untrustworthy model or prediction into a trustworthy one. In this work, we propose LIME, a novel explanation technique that explains the predictions of any classifier in an interpretable and faithful manner, by learning an interpretable model locally around the prediction. We further propose a method to explain models by presenting representative individual predictions and their explanations in a non-redundant way, framing the task as a submodular optimization problem. We demonstrate the flexibility of these methods by explaining different models for text (e.g. random forests) and image classification (e.g. neural networks). The usefulness of explanations is shown via novel experiments, both simulated and with human subjects. Our explanations empower users in various scenarios that require trust: deciding if one should trust a prediction, choosing between models, improving an untrustworthy classifier, and detecting why a classifier should not be trusted.
The κ coefficient is a popular descriptive statistic for summarizing an agreement table. It is sometimes desirable to combine some of the categories, for example, when categories are easily confused, and then calculate κ for the collapsed table. Since the categories of an agreement table are nominal and the order in which the categories of a table are listed is irrelevant, combining categories of an agreement table is identical to partitioning the categories in subsets.In this paper we prove that given a partition type of the categories, the overall κ-value of the original table is a weighted average of the κ-values of the collapsed tables corresponding to all partitions of that type. The weights are the denominators of the kappas of the subtables. An immediate consequence is that Cohen’s κ can be interpreted as a weighted average of the κ-values of the agreement tables corresponding to all non-trivial partitions.The κ-value of the 2×2 table that is obtained by combining all categories other than the one of current interest into a single “all others” category, reflects the reliability of the individual category. Since the overall κ-value is a weighted average of these 2×2 κ-values the category reliability indicates how a category contributes to the overall κ-value. It would be good practice to report both the overall κ-value and the category reliabilities of an agreement table.
The key idea behind active learning is that a machine learning algorithm can achieve greater accuracy with fewer labeled training instances if it is allowed to choose the data from which is learns. An active learner may ask queries in the form of unlabeled instances to be labeled by an oracle (e.g., a human annotator). Active learning is well-motivated in many modern machine learning problems, where unlabeled data may be abundant but labels are difficult, time-consuming, or expensive to obtain. This report provides a general introduction to active learning and a survey of the literature. This includes a discussion of the scenarios in which queries can be formulated, and an overview of the query strategy frameworks proposed in the literature to date. An analysis of the empirical and theoretical evidence for active learning, a summary of several problem setting variants, and a discussion of related topics in machine learning research are also presented.
A Framework for Understanding Unintended Consequences of Machine Learning
- H Suresh
- J V Guttag
Can Everyday AI be Ethical? Machine Learning Algorithm Fairness
- P Besse
- C Castets-Renard
- A Garivier
- J.-M Loubes
Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification
- J Buolamwini
- T Gebru
Movies fight detection dataset
- E B Nievas
- O D Suarez
- G B Garcia
- R Sukthankar
Movies fight detection dataset
- nievas