Publications

Decoding face recognition abilities in the human brain
Simon Faghel-Soubeyrand
Meike Ramon
Eva Bamps
Matteo Zoia
Jessica Woodhams
Anne-Raphaelle Richoz
Roberto Caldara
Frédéric Gosselin
Ian Charest
Why are some individuals better at recognising faces? Uncovering the neural mechanisms supporting face recognition ability has proven elusiv… (see more)e. To tackle this challenge, we used a multi-modal data-driven approach combining neuroimaging, computational modelling, and behavioural tests. We recorded the high-density electroencephalographic brain activity of individuals with extraordinary face recognition abilities—super-recognisers—and typical recognisers in response to diverse visual stimuli. Using multivariate pattern analyses, we decoded face recognition abilities from 1 second of brain activity with up to 80% accuracy. To better understand the mechanisms subtending this decoding, we compared computations in the brains of our participants with those in artificial neural network models of vision and semantics, as well as with those involved in human judgments of shape and meaning similarity. Compared to typical recognisers, we found stronger associations between early brain computations of super-recognisers and mid-level computations of vision models as well as shape similarity judgments. Moreover, we found stronger associations between late brain representations of super-recognisers and computations of the artificial semantic model as well as meaning similarity judgments. Overall, these results indicate that important individual variations in brain processing, including neural computations extending beyond purely visual processes, support differences in face recognition abilities. They provide the first empirical evidence for an association between semantic computations and face recognition abilities. We believe that such multi-modal data-driven approaches will likely play a critical role in further revealing the complex nature of idiosyncratic face recognition in the human brain. The ability to robustly recognise faces is crucial to our success as social beings. Yet, we still know little about the brain mechanisms allowing some individuals to excel at face recognition. This study builds on a sizeable neural dataset measuring the brain activity of individuals with extraordinary face recognition abilities—super-recognisers—to tackle this challenge. Using state-of-the-art computational methods, we show robust prediction of face recognition abilities in single individuals from a mere second of brain activity, and revealed specific brain computations supporting individual differences in face recognition ability. Doing so, we provide direct empirical evidence for an association between semantic computations and face recognition abilities in the human brain—a key component of prominent face recognition models.
2024-02-29
PNAS Nexus (published)
doi.org
Dissecting Deep RL with High Update Ratios: Combatting Value Overestimation and Divergence
Marcel Hussing
Claas Voelcker
Igor Gilitschenski
Amir-massoud Farahmand
Eric R. Eaton
We show that deep reinforcement learning can maintain its ability to learn without resetting network parameters in settings where the number… (see more) of gradient updates greatly exceeds the number of environment samples. Under such large update-to-data ratios, a recent study by Nikishin et al. (2022) suggested the emergence of a primacy bias , in which agents overfit early interactions and downplay later experience, impairing their ability to learn. In this work, we dissect the phenomena underlying the primacy bias. We inspect the early stages of training that ought to cause the failure to learn and find that a fundamental challenge is a long-standing acquaintance: value overestimation. Overinflated Q-values are found not only on out-of-distribution but also in-distribution data and can be traced to unseen action prediction propelled by optimizer momentum. We employ a simple unit-ball normalization that enables learning under large update ratios, show its efficacy on the widely used dm_control suite, and obtain strong performance on the challenging dog tasks, competitive with model-based approaches. Our results question, in parts, the prior explanation for sub-optimal learning due to overfitting on early data.
2024-02-29
arXiv (published)
doi.org
FedSwarm: An Adaptive Federated Learning Framework for Scalable AIoT
Haizhou Du
Chengdong Ni
Chaoqian Cheng
Qiao Xiang
X. T. Chen
Xue Liu
Federated learning (FL) is a key solution for datadriven the Artificial Intelligence of Things (AIoT). Although much progress has been made,… (see more) scalability remains a core challenge for real-world FL deployments. Existing solutions either suffer from accuracy loss or do not fully address the connectivity dynamicity of FL systems. In this article, we tackle the scalability issue with a novel, adaptive FL framework called FedSwarm, which improves system scalability for AIoT by deploying multiple collaborative edge servers. FedSwarm has two novel features: 1) adaptiveness on the number of local updates and 2) dynamicity of the synchronization between edge devices and edge servers. We formulate FedSwarm as a local update adaptation and perdevice dynamic server selection problem and prove FedSwarm‘s convergence bound. We further design a control mechanism consisting of a learning-based algorithm for collaboratively providing local update adaptation on the servers’ side and a bonus-based strategy for spurring dynamic per-device server selection on the devices’ side. Our extensive evaluation shows that FedSwarm significantly outperforms other studies with better scalability, lower energy consumption, and higher model accuracy.
2024-02-29
IEEE Internet of Things Journal (published)
doi.org
Gaussian-process-based Bayesian optimization for neurostimulation interventions in rats
Leo Choiniere
Rose Guay-Hottin
Rémi Picard
Guillaume Lajoie
Marco Bonizzato
Numa Dancause
2024-02-29
STAR Protocols (published)
doi.org
Generalization of deep learning models for hepatic steatosis grading using B-mode ultrasound images
Pedro Vianna
Yue Qi
Michael Chassé
Guy Wolf
Eugene Belilovsky
An Tang
Guy Cloutier
Grayscale ultrasound remains a key modality for screening of hepatic steatosis due to its non-invasiveness and availability. While neural ne… (see more)tworks have shown promise in this field, their main drawback lies in their inability to generalize to diverse real-world settings. Variations in equipment, acquisition parameters, or population significantly affect model performance. Test-time adaptation, an unsupervised domain adaptation technique, overcomes these limitations by adjusting trained models during inference. Our retrospective study used two datasets collected in separate populations, with different scanners and protocols. We propose an adaptation method, using test-time batch normalization to selectively adjust BatchNorm layers based on test data for predicting steatosis grades. Comparing the non-adapted and adapted models, the mean absolute error (± standard deviation) in grading four severities of steatosis decreased from 0.92 ± 0.21 to 0.64 ± 0.22 . Specifically, for detection of steatosis the area under the curve increased from 0.76 ± 0.05 to 0.95 ± 0.02 when using the adapted model. Adapted models show promising results in improving performance compared to base models when testing data differ significantly from training data. Results suggest that the proposed method effectively addresses domain shift in diagnosing fatty liver using ultrasound images, reducing risks associated with deploying trained models.
2024-02-29
The Journal of the Acoustical Society of America (published)
doi.org
Heterogeneous ensemble prediction model of CO emission concentration in municipal solid waste incineration process using virtual data and real data hybrid-driven
Runyu Zhang
Jian Tang
Heng Xia
Jiakun Chen
Wen Yu
JunFei Qiao
2024-02-29
Journal of Cleaner Production (published)
doi.org
Implications of conscious AI in primary healthcare
Dorsai Ranjbari
Samira Abbasgholizadeh Rahimi
The conversation about consciousness of artificial intelligence (AI) is an ongoing topic since 1950s. Despite the numerous applications of A… (see more)I identified in healthcare and primary healthcare, little is known about how a conscious AI would reshape its use in this domain. While there is a wide range of ideas as to whether AI can or cannot possess consciousness, a prevailing theme in all arguments is uncertainty. Given this uncertainty and the high stakes associated with the use of AI in primary healthcare, it is imperative to be prepared for all scenarios including conscious AI systems being used for medical diagnosis, shared decision-making and resource management in the future. This commentary serves as an overview of some of the pertinent evidence supporting the use of AI in primary healthcare and proposes ideas as to how consciousnesses of AI can support or further complicate these applications. Given the scarcity of evidence on the association between consciousness of AI and its current state of use in primary healthcare, our commentary identifies some directions for future research in this area including assessing patients’, healthcare workers’ and policy-makers’ attitudes towards consciousness of AI systems in primary healthcare settings.
2024-02-29
Family Medicine and Community Health (published)
doi.org
Iterative Graph Self-Distillation
Hanlin Zhang
Shuai Lin
Weiyang Liu
Pan Zhou
Jian Tang
Xiaodan Liang
Eric P. Xing
Recently, there has been increasing interest in the challenge of how to discriminatively vectorize graphs. To address this, we propose a met… (see more)hod called Iterative Graph Self-Distillation (IGSD) which learns graph-level representation in an unsupervised manner through instance discrimination using a self-supervised contrastive learning approach. IGSD involves a teacher-student distillation process that uses graph diffusion augmentations and constructs the teacher model using an exponential moving average of the student model. The intuition behind IGSD is to predict the teacher network representation of the graph pairs under different augmented views. As a natural extension, we also apply IGSD to semi-supervised scenarios by jointly regularizing the network with both supervised and self-supervised contrastive loss. Finally, we show that fine-tuning the IGSD-trained models with self-training can further improve graph representation learning. Empirically, we achieve significant and consistent performance gain on various graph datasets in both unsupervised and semi-supervised settings, which well validates the superiority of IGSD.
2024-02-29
IEEE Transactions on Knowledge and Data Engineering (published)
doi.org
openreview.net
Neural network prediction of the effect of thermomechanical controlled processing on mechanical properties
Sushant Sinha
Denzel Guye
Xiaoping Ma
Kashif Rehman
Stephen Yue
Narges Armanfard
2024-02-29
Machine Learning with Applications (published)
doi.org
Novel community data in ecology-properties and prospects.
Florian Hartig
Nerea Abrego
Alex Bush
Jonathan M. Chase
G. Guillera‐Arroita
M. Leibold
Otso T. Ovaskainen
Loïc Pellissier
Maximilian Pichler
Giovanni Poggiato
Laura J. Pollock
Sara Si-moussi
Wilfried Thuiller
Duarte S Viana
D. Warton
Damaris Zurell
Douglas W. Yu
2024-02-29
Trends in Ecology & Evolution (published)
doi.org
arxiv.org
Reply to: Model uncertainty obscures major driver of soil carbon
Feng Tao
Benjamin Z. Houlton
Serita D. Frey
Johannes Lehmann
Stefano Manzoni
Yuanyuan Huang
Lifen Jiang
Umakant Mishra
Bruce A. Hungate
Michael W. I. Schmidt
Markus Reichstein
Nuno Carvalhais
Philippe Ciais
Ying-Ping Wang
Bernhard Ahrens
Gustaf Hugelius
Toby Dylan Hocking
Xingjie Lu
Zheng Shi
Kostiantyn Viatkin … (see 15 more)
K. Viatkin
Ronald Vargas
Yusuf Yigini
Christian Omuto
Ashish A. Malik
Guillermo Peralta
Rosa Cuevas-Corona
Luciano E. Di Paolo
Isabel Luotto
Cuijuan Liao
Yi-Shuang Liang
Yixin Liang
Vinisa S. Saynes
Xiaomeng Huang
Yiqi Luo
2024-02-29
Nature (published)
doi.org
Revisiting Dynamic Evaluation: Online Adaptation for Large Language Models
Amal Rannen-Triki
Jörg Bornschein
Razvan Pascanu
Marcus Hutter
Andr'as Gyorgy
Alexandre Galashov
Yee Whye Teh
Michalis K. Titsias
We consider the problem of online fine tuning the parameters of a language model at test time, also known as dynamic evaluation. While it is… (see more) generally known that this approach improves the overall predictive performance, especially when considering distributional shift between training and evaluation data, we here emphasize the perspective that online adaptation turns parameters into temporally changing states and provides a form of context-length extension with memory in weights, more in line with the concept of memory in neuroscience. We pay particular attention to the speed of adaptation (in terms of sample efficiency),sensitivity to the overall distributional drift, and the computational overhead for performing gradient computations and parameter updates. Our empirical study provides insights on when online adaptation is particularly interesting. We highlight that with online adaptation the conceptual distinction between in-context learning and fine tuning blurs: both are methods to condition the model on previously observed tokens.
2024-02-29
arXiv (published)
doi.org
arxiv.org