Research Papers

Face morphing attacks can compromise Face Recognition System (FRS) by exploiting their vulnerability. Face Morphing Attack Detection (MAD) techniques have been developed in recent past to deter such attacks and mitigate risks from morphing attacks. MAD algorithms, as any other algorithms should treat the images of subjects from different ethnic origins in an equal manner and provide non-discriminatory results. While the promising MAD algorithms are tested for robustness, there is no study comprehensively bench-marking their behaviour against various ethnicities. In this paper, we study and present a comprehensive analysis of algorithmic fairness of the existing Single image-based Morph Attack Detection (S-MAD) algorithms. We attempt to better understand the influence of ethnic bias on MAD algorithms and to this extent, we study the performance of MAD algorithms on a newly created dataset consisting of four different ethnic groups. With Extensive experiments using six different S-MAD techniques, we first present benchmark of detection performance and then measure the quantitative value of the algorithmic fairness for each of them using Fairness Discrepancy Rate (FDR). The results indicate the lack of fairness on all six different S-MAD methods when trained and tested on different ethnic groups suggesting the need for reliable MAD approaches to mitigate the algorithmic bias.

Advances in machine learning (ML) technologies have greatly improved Artificial Intelligence (AI) systems. As a result, AI systems have become ubiquitous, with their application prevalent in virtually all sectors. However, AI systems have prompted ethical concerns, especially as their usage crosses boundaries in sensitive areas such as healthcare, transportation, and security. As a result, users are calling for better AI governance practices in ethical AI systems. Therefore, AI development methods are encouraged to foster these practices. This research analyzes the ECCOLA method for developing ethical and trustworthy AI systems to determine if it enables AI governance in development processes through ethical practices. The results demonstrate that while ECCOLA fully facilitates AI governance in corporate governance practices in all its processes, some of its practices do not fully foster data governance and information governance practices. This indicates that the method can be further improved.

Principles of fairness and solidarity in AI ethics regularly overlap, creating obscurity in practice: acting in accordance with one can appear indistinguishable from deciding according to the rules of the other. However, there exist irregular cases where the two concepts split, and so reveal their disparate meanings and uses. This paper explores two cases in AI medical ethics, one that is irregular and the other more conventional, to fully distinguish fairness and solidarity. Then the distinction is applied to the frequently cited COMPAS versus ProPublica dispute in judicial ethics. The application provides a broader model for settling contemporary and topical debates about fairness and solidarity. It also implies a deeper and disorienting truth about AI ethics principles and their justification.

Algorithmic Bias can be due to bias in the training data or issues with the algorithm itself. These algorithmic issues typically relate to problems with model capacity and regularisation. This underestimation bias may arise because the model has been optimised for good generalisation accuracy without any explicit consideration of bias or fairness. In a sense, we should not be surprised that a model might be biased when it hasn't been "asked" not to be. In this paper, we consider including bias (underestimation) as an additional criterion in model training. We present a multi-objective optimisation strategy using Pareto Simulated Annealing that optimise for both balanced accuracy and underestimation. We demonstrate the effectiveness of this strategy on one synthetic and two real-world datasets.

For AI technology to fulfill its full promises, we must have effective means to ensure Responsible AI behavior and curtail potential irresponsible use, e.g., in areas of privacy protection, human autonomy, robustness, and prevention of biases and discrimination in automated decision making. Recent literature in the field has identified serious shortcomings of narrow technology focused and formalism-oriented research and has proposed an interdisciplinary approach that brings the social context into the scope of study. In this paper, we take a sociotechnical approach to propose a more expansive framework of thinking about the Responsible AI challenges in both technical and social context. Effective solutions need to bridge the gap between a technical system with the social system that it will be deployed to. To this end, we propose human agency and regulation as main mechanisms of intervention and propose a decentralized computational infrastructure, or a set of public utilities, as the computational means to bridge this gap. A decentralized infrastructure is uniquely suited for meeting this challenge and enable technical solutions and social institutions in a mutually reinforcing dynamic to achieve Responsible AI goals. Our approach is novel in its sociotechnical approach and its aim in tackling the structural issues that cannot be solved within the narrow confines of AI technical research. We then explore possible features of the proposed infrastructure and discuss how it may help solve example problems recently studied in the field.

Explainable AI (XAI) is paramount in industry-grade AI; however existing methods fail to address this necessity, in part due to a lack of standardisation of explainability methods. The purpose of this paper is to offer a perspective on the current state of the area of explainability, and to provide novel definitions for Explainability and Interpretability to begin standardising this area of research. To do so, we provide an overview of the literature on explainability, and of the existing methods that are already implemented. Finally, we offer a tentative taxonomy of the different explainability methods, opening the door to future research.

Stuttering Speech Disfluency Prediction using Explainable Attribution Vectors of Facial Muscle Movements

Arun Das, Jeffrey Mock, Henry Chacon +3

Speech disorders such as stuttering disrupt the normal fluency of speech by involuntary repetitions, prolongations and blocking of sounds and syllables. In addition to these disruptions to speech fluency, most adults who stutter (AWS) also experience numerous observable secondary behaviors before, during, and after a stuttering moment, often involving the facial muscles. Recent studies have explored automatic detection of stuttering using Artificial Intelligence (AI) based algorithm from respiratory rate, audio, etc. during speech utterance. However, most methods require controlled environments and/or invasive wearable sensors, and are unable explain why a decision (fluent vs stuttered) was made. We hypothesize that pre-speech facial activity in AWS, which can be captured non-invasively, contains enough information to accurately classify the upcoming utterance as either fluent or stuttered. Towards this end, this paper proposes a novel explainable AI (XAI) assisted convolutional neural network (CNN) classifier to predict near future stuttering by learning temporal facial muscle movement patterns of AWS and explains the important facial muscles and actions involved. Statistical analyses reveal significantly high prevalence of cheek muscles (p<0.005) and lip muscles (p<0.005) to predict stuttering and shows a behavior conducive of arousal and anticipation to speak. The temporal study of these upper and lower facial muscles may facilitate early detection of stuttering, promote automated assessment of stuttering and have application in behavioral therapies by providing automatic non-invasive feedback in realtime.

Instances of Artificial Intelligence (AI) systems failing to deliver consistent, satisfactory performance are legion. We investigate why AI failures occur. We address only a narrow subset of the broader field of AI Safety. We focus on AI failures on account of flaws in conceptualization, design and deployment. Other AI Safety issues like trade-offs between privacy and security or convenience, bad actors hacking into AI systems to create mayhem or bad actors deploying AI for purposes harmful to humanity and are out of scope of our discussion. We find that AI systems fail on account of omission and commission errors in the design of the AI system, as well as upon failure to develop an appropriate interpretation of input information. Moreover, even when there is no significant flaw in the AI software, an AI system may fail because the hardware is incapable of robust performance across environments. Finally an AI system is quite likely to fail in situations where, in effect, it is called upon to deliver moral judgments -- a capability AI does not possess. We observe certain trade-offs in measures to mitigate a subset of AI failures and provide some recommendations.

Explaining How Deep Neural Networks Forget by Deep Visualization

Giang Nguyen, Shuan Chen, Tae Joon Jun +1

Explaining the behaviors of deep neural networks, usually considered as black boxes, is critical especially when they are now being adopted over diverse aspects of human life. Taking the advantages of interpretable machine learning (interpretable ML), this paper proposes a novel tool called Catastrophic Forgetting Dissector (or CFD) to explain catastrophic forgetting in continual learning settings. We also introduce a new method called Critical Freezing based on the observations of our tool. Experiments on ResNet articulate how catastrophic forgetting happens, particularly showing which components of this famous network are forgetting. Our new continual learning algorithm defeats various recent techniques by a significant margin, proving the capability of the investigation. Critical freezing not only attacks catastrophic forgetting but also exposes explainability.

Current techniques for explainable AI have been applied with some success to image processing. The recent rise of research in video processing has called for similar work n deconstructing and explaining spatio-temporal models. While many techniques are designed for 2D convolutional models, others are inherently applicable to any input domain. One such body of work, deep Taylor decomposition, propagates relevance from the model output distributively onto its input and thus is not restricted to image processing models. However, by exploiting a simple technique that removes motion information, we show that it is not the case that this technique is effective as-is for representing relevance in non-image tasks. We instead propose a discriminative method that produces a naïve representation of both the spatial and temporal relevance of a frame as two separate objects. This new discriminative relevance model exposes relevance in the frame attributed to motion, that was previously ambiguous in the original explanation. We observe the effectiveness of this technique on a range of samples from the UCF-101 action recognition dataset, two of which are demonstrated in this paper.

Autonomous agents trained via reinforcement learning present numerous safety concerns: reward hacking, negative side effects, and unsafe exploration, among others. In the context of near-future autonomous agents, operating in environments where humans understand the existing dangers, human involvement in the learning process has proved a promising approach to AI Safety. Here we demonstrate that a precise framework for learning from human input, loosely inspired by the way humans parent children, solves a broad class of safety problems in this context. We show that our Parenting algorithm solves these problems in the relevant AI Safety gridworlds of Leike et al. (2017), that an agent can learn to outperform its parent as it "matures", and that policies learnt through Parenting are generalisable to new environments.

We present the Membership Inference Test (MINT) Demo 2, a framework designed to improve transparency in machine learning training processes. MINT is a technique for experimentally determining whether specific data were used during machine learning model training. We establish the theoretical framework and propose multiple architectures for MINT depending on the amount of information known about the models that are being audited. Experimental results using a popular face recognition model, 4 state-of-the-art LLMs, and multiple, diverse, and large-scale public image and text databases achieve promising accuracy levels in the detection of training data of up to 90%. Building on these results, we introduce a comprehensive web platform1 that expands these capabilities to image and text modalities. The platform integrates a diverse technological stack, including MINT, aMINT, and gMINT, allowing users to audit a wide range of models. This demonstrator aims to promote AI transparency and provides a practical tool to foster compliance with emerging AI regulations.