-
Everyday algorithm auditing: Understanding the power of everyday users in surfacing harmful algorithmic behaviors
Authors:
Hong Shen,
Alicia DeVos,
Motahhare Eslami,
Kenneth Holstein
Abstract:
A growing body of literature has proposed formal approaches to audit algorithmic systems for biased and harmful behaviors. While formal auditing approaches have been greatly impactful, they often suffer major blindspots, with critical issues surfacing only in the context of everyday use once systems are deployed. Recent years have seen many cases in which everyday users of algorithmic systems dete…
▽ More
A growing body of literature has proposed formal approaches to audit algorithmic systems for biased and harmful behaviors. While formal auditing approaches have been greatly impactful, they often suffer major blindspots, with critical issues surfacing only in the context of everyday use once systems are deployed. Recent years have seen many cases in which everyday users of algorithmic systems detect and raise awareness about harmful behaviors that they encounter in the course of their everyday interactions with these systems. However, to date little academic attention has been granted to these bottom-up, user-driven auditing processes. In this paper, we propose and explore the concept of everyday algorithm auditing, a process in which users detect, understand, and interrogate problematic machine behaviors via their day-to-day interactions with algorithmic systems. We argue that everyday users are powerful in surfacing problematic machine behaviors that may elude detection via more centrally-organized forms of auditing, regardless of users' knowledge about the underlying algorithms. We analyze several real-world cases of everyday algorithm auditing, drawing lessons from these cases for the design of future platforms and tools that facilitate such auditing behaviors. Finally, we discuss work that lies ahead, toward bridging the gaps between formal auditing approaches and the organic auditing behaviors that emerge in everyday use of algorithmic systems.
△ Less
Submitted 24 August, 2021; v1 submitted 6 May, 2021;
originally announced May 2021.
-
Model-Agnostic Learning to Meta-Learn
Authors:
Arnout Devos,
Yatin Dandi
Abstract:
In this paper, we propose a learning algorithm that enables a model to quickly exploit commonalities among related tasks from an unseen task distribution, before quickly adapting to specific tasks from that same distribution. We investigate how learning with different task distributions can first improve adaptability by meta-finetuning on related tasks before improving goal task generalization wit…
▽ More
In this paper, we propose a learning algorithm that enables a model to quickly exploit commonalities among related tasks from an unseen task distribution, before quickly adapting to specific tasks from that same distribution. We investigate how learning with different task distributions can first improve adaptability by meta-finetuning on related tasks before improving goal task generalization with finetuning. Synthetic regression experiments validate the intuition that learning to meta-learn improves adaptability and consecutively generalization. Experiments on more complex image classification, continual regression, and reinforcement learning tasks demonstrate that learning to meta-learn generally improves task-specific adaptation. The methodology, setup, and hypotheses in this proposal were positively evaluated by peer review before conclusive experiments were carried out.
△ Less
Submitted 19 July, 2021; v1 submitted 4 December, 2020;
originally announced December 2020.
-
Fano Resonance between Stokes and Anti-Stokes Brillouin Scattering
Authors:
KwanTo Lai,
Daniel Finkelstein-Shapiro,
Arnaud Devos,
Pierre-Adrien Mante
Abstract:
In recent years, the manipulation of Fano resonances in the time domain has unlocked deep insights into a broad spectrum of systems' coherent dynamics. Here, inelastic scattering of light with coherent acoustic phonons is harnessed to achieve complex Fano resonances. The sudden change of phonon momentum during reflection leads to a transition from anti-Stokes to Stokes light scattering, producing…
▽ More
In recent years, the manipulation of Fano resonances in the time domain has unlocked deep insights into a broad spectrum of systems' coherent dynamics. Here, inelastic scattering of light with coherent acoustic phonons is harnessed to achieve complex Fano resonances. The sudden change of phonon momentum during reflection leads to a transition from anti-Stokes to Stokes light scattering, producing two different resonances that interfere in the measurement process. We highlight the conditions necessary to achieve such interference, revealing an underlying symmetry between photons and phonons, and verify the theory experimentally. Then, we demonstrate the possibility to characterize energy and coherence losses at rough interfaces, thus providing a mechanism for nondestructive testing of interface quality. Our results describe numerous unexplained observations in ultrafast acoustics and can be generalized to the scattering of light with any waves.
△ Less
Submitted 21 October, 2020;
originally announced October 2020.
-
Self-Supervised Prototypical Transfer Learning for Few-Shot Classification
Authors:
Carlos Medina,
Arnout Devos,
Matthias Grossglauser
Abstract:
Most approaches in few-shot learning rely on costly annotated data related to the goal task domain during (pre-)training. Recently, unsupervised meta-learning methods have exchanged the annotation requirement for a reduction in few-shot classification performance. Simultaneously, in settings with realistic domain shift, common transfer learning has been shown to outperform supervised meta-learning…
▽ More
Most approaches in few-shot learning rely on costly annotated data related to the goal task domain during (pre-)training. Recently, unsupervised meta-learning methods have exchanged the annotation requirement for a reduction in few-shot classification performance. Simultaneously, in settings with realistic domain shift, common transfer learning has been shown to outperform supervised meta-learning. Building on these insights and on advances in self-supervised learning, we propose a transfer learning approach which constructs a metric embedding that clusters unlabeled prototypical samples and their augmentations closely together. This pre-trained embedding is a starting point for few-shot classification by summarizing class clusters and fine-tuning. We demonstrate that our self-supervised prototypical transfer learning approach ProtoTransfer outperforms state-of-the-art unsupervised meta-learning methods on few-shot tasks from the mini-ImageNet dataset. In few-shot experiments with domain shift, our approach even has comparable performance to supervised methods, but requires orders of magnitude fewer labels.
△ Less
Submitted 19 June, 2020;
originally announced June 2020.
-
Revisiting Few-Shot Learning for Facial Expression Recognition
Authors:
Anca-Nicoleta Ciubotaru,
Arnout Devos,
Behzad Bozorgtabar,
Jean-Philippe Thiran,
Maria Gabrani
Abstract:
Most of the existing deep neural nets on automatic facial expression recognition focus on a set of predefined emotion classes, where the amount of training data has the biggest impact on performance. However, in the standard setting over-parameterised neural networks are not amenable for learning from few samples as they can quickly over-fit. In addition, these approaches do not have such a strong…
▽ More
Most of the existing deep neural nets on automatic facial expression recognition focus on a set of predefined emotion classes, where the amount of training data has the biggest impact on performance. However, in the standard setting over-parameterised neural networks are not amenable for learning from few samples as they can quickly over-fit. In addition, these approaches do not have such a strong generalisation ability to identify a new category, where the data of each category is too limited and significant variations exist in the expression within the same semantic category. We embrace these challenges and formulate the problem as a low-shot learning, where once the base classifier is deployed, it must rapidly adapt to recognise novel classes using a few samples. In this paper, we revisit and compare existing few-shot learning methods for the low-shot facial expression recognition in terms of their generalisation ability via episode-training. In particular, we extend our analysis on the cross-domain generalisation, where training and test tasks are not drawn from the same distribution. We demonstrate the efficacy of low-shot learning methods through extensive experiments.
△ Less
Submitted 11 December, 2019; v1 submitted 5 December, 2019;
originally announced December 2019.
-
Properties of Nanocrystalline Silicon Probed by Optomechanics
Authors:
Daniel Navarro-Urrios,
Martin F. Colombano,
Jeremie Maire,
Emigdio Chavez-Angel,
Guillermo Arregui,
Nestor E. Capuj,
Arnaud Devos,
Amadeu Griol,
Laurent Bellieres,
Alejandro Martinez,
Kestutis Grigoras,
Teija Hakkinen,
Jaakko Saarilahti,
Tapani Makkonen,
Clivia M. Sotomayor-Torres,
Jouni Ahopelto
Abstract:
Nanocrystalline materials exhibit properties that can differ substantially from those of their single crystal counterparts. As such, they provide ways to enhance and optimise their functionality for devices and applications. Here we report on the optical, mechanical and thermal properties of nanocrystalline silicon probed by means of optomechanical nanobeams to extract information of the dynamics…
▽ More
Nanocrystalline materials exhibit properties that can differ substantially from those of their single crystal counterparts. As such, they provide ways to enhance and optimise their functionality for devices and applications. Here we report on the optical, mechanical and thermal properties of nanocrystalline silicon probed by means of optomechanical nanobeams to extract information of the dynamics of optical absorption, mechanical losses, heat generation and dissipation. The optomechanical nanobeams are fabricated using nanocrystalline films prepared by annealing amorphous silicon layers at different temperatures. The resulting crystallite sizes and the stress in the films can be controlled by the annealing temperature and time and, consequently, the properties of the films can be tuned relatively freely, as demonstrated here by means of electron microscopy and Raman scattering. We show that the nanocrystallite size and the volume fraction of the grain boundaries play a key role in the dissipation rates through non-linear optical and thermal processes. Promising optical (13000) and mechanical (1700) quality factors were found in the optomechanical cavity realised in the nanocrystalline Si resulting from annealing at 950 C. The enhanced absorption and recombination rates via the intra-gap states and the reduced thermal conductivity boost the potential to exploit these non-linear effects in applications, including NEMS, phonon lasing and chaos-based devices.
△ Less
Submitted 2 September, 2020; v1 submitted 9 July, 2019;
originally announced July 2019.
-
Regression Networks for Meta-Learning Few-Shot Classification
Authors:
Arnout Devos,
Matthias Grossglauser
Abstract:
We propose regression networks for the problem of few-shot classification, where a classifier must generalize to new classes not seen in the training set, given only a small number of examples of each class. In high dimensional embedding spaces the direction of data generally contains richer information than magnitude. Next to this, state-of-the-art few-shot metric methods that compare distances w…
▽ More
We propose regression networks for the problem of few-shot classification, where a classifier must generalize to new classes not seen in the training set, given only a small number of examples of each class. In high dimensional embedding spaces the direction of data generally contains richer information than magnitude. Next to this, state-of-the-art few-shot metric methods that compare distances with aggregated class representations, have shown superior performance. Combining these two insights, we propose to meta-learn classification of embedded points by regressing the closest approximation in every class subspace while using the regression error as a distance metric. Similarly to recent approaches for few-shot learning, regression networks reflect a simple inductive bias that is beneficial in this limited-data regime and they achieve excellent results, especially when more aggregate class representations can be formed with multiple shots.
△ Less
Submitted 18 June, 2020; v1 submitted 31 May, 2019;
originally announced May 2019.
-
Strong coronal channelling and interplanetary evolution of a solar storm up to Earth and Mars
Authors:
Christian Möstl,
Tanja Rollett,
Rudy A. Frahm,
Ying D. Liu,
David M. Long,
Robin C. Colaninno,
Martin A. Reiss,
Manuela Temmer,
Charles J. Farrugia,
Arik Posner,
Mateja Dumbović,
Miho Janvier,
Pascal Démoulin,
Peter Boakes,
Andy Devos,
Emil Kraaikamp,
Mona L. Mays,
Bojan Vrsnak
Abstract:
The severe geomagnetic effects of solar storms or coronal mass ejections (CMEs) are to a large degree determined by their propagation direction with respect to Earth. There is a lack of understanding of the processes that determine their non-radial propagation. Here we present a synthesis of data from seven different space missions of a fast CME, which originated in an active region near the disk…
▽ More
The severe geomagnetic effects of solar storms or coronal mass ejections (CMEs) are to a large degree determined by their propagation direction with respect to Earth. There is a lack of understanding of the processes that determine their non-radial propagation. Here we present a synthesis of data from seven different space missions of a fast CME, which originated in an active region near the disk centre and, hence, a significant geomagnetic impact was forecasted. However, the CME is demonstrated to be channelled during eruption into a direction + 37+/-10 degree (longitude) away from its source region, leading only to minimal geomagnetic effects. In situ observations near Earth and Mars confirm the channelled CME motion, and are consistent with an ellipse shape of the CME-driven shock provided by the new Ellipse Evolution model, presented here. The results enhance our understanding of CME propagation and shape, which can help to improve space weather forecasts.
△ Less
Submitted 10 June, 2015; v1 submitted 9 June, 2015;
originally announced June 2015.
-
Geoeffectiveness of Coronal Mass Ejections in the SOHO era
Authors:
Mateja Dumbovic,
Andy Devos,
Bojan Vrsnak,
Davor Sudar,
Luciano Rodriguez,
Domagoj Ruzdjak,
Kristoffer Leer,
Susanne Vennerstrom,
Astrid Veronig
Abstract:
The main objective of the study is to determine the probability distributions of the geomagnetic Dst index as a function of the coronal mass ejection (CME) and solar flare parameters for the purpose of establishing a probabilistic forecast tool for the geomagnetic storm intensity. Several CME and flare parameters as well as the effect of successive-CME occurrence in changing the probability for a…
▽ More
The main objective of the study is to determine the probability distributions of the geomagnetic Dst index as a function of the coronal mass ejection (CME) and solar flare parameters for the purpose of establishing a probabilistic forecast tool for the geomagnetic storm intensity. Several CME and flare parameters as well as the effect of successive-CME occurrence in changing the probability for a certain range of Dst index values, were examined. The results confirm some of already known relationships between remotely-observed properties of solar eruptive events and geomagnetic storms, namely the importance of initial CME speed, apparent width, source position, and the associated solar flare class. In this paper we quantify these relationships in a form to be used for space weather forecasting in future. The results of the statistical study are employed to construct an empirical statistical model for predicting the probability of the geomagnetic storm intensity based on remote solar observations of CMEs and flares.
△ Less
Submitted 13 October, 2014;
originally announced October 2014.
-
Hypersound damping in vitreous silica measured by picosecond acoustics
Authors:
A. Devos,
M. Foret,
S. Ayrinhac,
P. Emery,
B. Rufflé
Abstract:
The attenuation of longitudinal acoustic phonons up to frequencies nearing 250 GHz is measured in vitreous silica with a picosecond optical technique. Taking advantage of interferences on the probe beam, difficulties encountered in early pioneering experiments are alleviated. Sound damping at 250 GHz and room temperature is consistent with relaxation dominated by anharmonic interactions with the…
▽ More
The attenuation of longitudinal acoustic phonons up to frequencies nearing 250 GHz is measured in vitreous silica with a picosecond optical technique. Taking advantage of interferences on the probe beam, difficulties encountered in early pioneering experiments are alleviated. Sound damping at 250 GHz and room temperature is consistent with relaxation dominated by anharmonic interactions with the thermal bath, extending optical Brillouin scattering data. Our result is at variance with claims of a recent deep-UV experiment which reported a rapid damping increase beyond 100 GHz. A comprehensive picture of the frequency dependence of sound attenuation in $v$-SiO$_2$ can be proposed.
△ Less
Submitted 6 March, 2008;
originally announced March 2008.
