Publications

Training Compute-Optimal Vision Transformers for Brain Encoding
Sana Ahmadi
Fraçois Paugam
Tristan Glatard
Lune P Bellec
The optimal training of a vision transformer for brain encoding depends on three factors: model size, data size, and computational resources… (voir plus). This study investigates these three pillars, focusing on the effects of data scaling, model scaling, and high-performance computing on brain encoding results. Using VideoGPT to extract efficient spatiotemporal features from videos and training a Ridge model to predict brain activity based on these features, we conducted benchmark experiments with varying data sizes (10k, 100k, 1M, 6M) and different model configurations of GPT-2, including hidden layer dimensions, number of layers, and number of attention heads. We also evaluated the effects of training models with 32-bit vs 16-bit floating point representations. Our results demonstrate that increasing the hidden layer dimensions significantly improves brain encoding performance, as evidenced by higher Pearson correlation coefficients across all subjects. In contrast, the number of attention heads does not have a significant effect on the encoding results. Additionally, increasing the number of layers shows some improvement in brain encoding correlations, but the trend is not as consistent as that observed with hidden layer dimensions. The data scaling results show that larger training datasets lead to improved brain encoding performance, with the highest Pearson correlation coefficients observed for the largest dataset size (6M). These findings highlight that the effects of data scaling are more significant compared to model scaling in enhancing brain encoding performance. Furthermore, we explored the impact of floating-point precision by comparing 32-bit and 16-bit representations. Training with 16-bit precision yielded the same brain encoding accuracy as 32-bit, while reducing training time by 1.17 times, demonstrating its efficiency for high-performance computing tasks.
2024-10-16
ArXiv (prépublication)
doi.org
arxiv.org
BlabberSeg: Real-Time Embedded Open-Vocabulary Aerial Segmentation
Haechan Mark Bong
Ricardo de Azambuja
Giovanni Beltrame
Real-time aerial image segmentation plays an important role in the environmental perception of Uncrewed Aerial Vehicles (UAVs). We introduce… (voir plus) BlabberSeg, an optimized Vision-Language Model built on CLIPSeg for on-board, real-time processing of aerial images by UAVs. BlabberSeg improves the efficiency of CLIPSeg by reusing prompt and model features, reducing computational overhead while achieving real-time open-vocabulary aerial segmentation. We validated BlabberSeg in a safe landing scenario using the Dynamic Open-Vocabulary Enhanced SafE-Landing with Intelligence (DOVESEI) framework, which uses visual servoing and open-vocabulary segmentation. BlabberSeg reduces computational costs significantly, with a speed increase of 927.41% (16.78 Hz) on a NVIDIA Jetson Orin AGX (64GB) compared with the original CLIPSeg (1.81Hz), achieving real-time aerial segmentation with negligible loss in accuracy (2.1% as the ratio of the correctly segmented area with respect to CLIPSeg). BlabberSeg's source code is open and available online.
2024-10-15
ArXiv (prépublication)
doi.org
arxiv.org
The Non-Local Model Merging Problem: Permutation Symmetries and Variance Collapse
Ekansh Sharma
Daniel M. Roy
Gintare Karolina Dziugaite
2024-10-15
ArXiv (prépublication)
doi.org
arxiv.org
WorldCuisines: A Massive-Scale Benchmark for Multilingual and Multicultural Visual Question Answering on Global Cuisines
Genta Indra Winata
Frederikus Hudi
Patrick Amadeus Irawan
David Anugraha
Rifki Afina Putri
Yutong Wang
Adam Nohejl
Ubaidillah Ariq Prathama
Nedjma OUSIDHOUM
Afifa Amriani
Anar Rzayev
Anirban Das
Ashmari Pramodya
Aulia Adila
Bryan Wilie
Candy Olivia Mawalim
Ching Lam Cheng
Daud Abolade
Emmanuele Chersoni
Enrico Santus … (voir 31 de plus)
Fariz Ikhwantri
Garry Kuwanto
Hanyang Zhao
Haryo Akbarianto Wibowo
Holy Lovenia
Jan Christian Blaise Cruz
Jan Wira Gotama Putra
Junho Myung
Lucky Susanto
Maria Angelica Riera Machin
Marina Zhukova
Michael Anugraha
Muhammad Farid Adilazuarda
Natasha Santosa
Peerat Limkonchotiwat
Raj Dabre
Rio Alexander Audino
Samuel Cahyawijaya
Shi-Xiong Zhang
Stephanie Yulia Salim
Yi Zhou
Yinxuan Gui
David Ifeoluwa Adelani
En-Shiun Annie Lee
Shogo Okada
Ayu Purwarianti
Alham Fikri Aji
Taro Watanabe
Derry Tanti Wijaya
Alice Oh
Chong-Wah Ngo
2024-10-15
ArXiv (prépublication)
doi.org
arxiv.org
Adversarial Bounding Boxes Generation (ABBG) Attack against Visual Object Trackers
Fatemeh Nourilenjan Nokabadi
Jean-François Lalonde
Christian Gagné
Adversarial perturbations aim to deceive neural networks into predicting inaccurate results. For visual object trackers, adversarial attacks… (voir plus) have been developed to generate perturbations by manipulating the outputs. However, transformer trackers predict a specific bounding box instead of an object candidate list, which limits the applicability of many existing attack scenarios. To address this issue, we present a novel white-box approach to attack visual object trackers with transformer backbones using only one bounding box. From the tracker predicted bounding box, we generate a list of adversarial bounding boxes and compute the adversarial loss for those bounding boxes. Experimental results demonstrate that our simple yet effective attack outperforms existing attacks against several robust transformer trackers, including TransT-M, ROMTrack, and MixFormer, on popular benchmark tracking datasets such as GOT-10k, UAV123, and VOT2022STS.
2024-10-14
NeurIPS.cc/2024/Workshop/AdvML-Frontiers (publié)
doi.org
openreview.net
Comparative evaluation of methodologies for estimating the effectiveness of non-pharmaceutical interventions in the context of COVID-19: a simulation study
Iris Ganser
Juliette Paireau
David L Buckeridge
Simon Cauchemez
Rodolphe Thiébaut
M. Prague
2024-10-14
medRxiv (prépublication)
doi.org
Learning to Forget using Hypernetworks
Jose Miguel Lara Rangel
Usman Anwar
Stefan Schoepf
Jack Foster
David M. Krueger
Machine unlearning is gaining increasing attention as a way to remove adversarial data poisoning attacks from already trained models and to … (voir plus)comply with privacy and AI regulations. The objective is to unlearn the effect of undesired data from a trained model while maintaining performance on the remaining data. This paper introduces HyperForget, a novel machine unlearning framework that leverages hypernetworks– neural networks that generate parameters for other networks– to dynamically sample models that lack knowledge of targeted data while preserving essential capabilities. Leveraging diffusion models, we implement two Diffusion HyperForget Networks and used them to sample unlearned models in Proof-of-Concept experiments. The unlearned models obtained zero accuracy on the forget set, while preserving good accuracy on the retain sets, highlighting the potential of HyperForget for dynamic targeted data removal and a promising direction for developing adaptive machine unlearning algorithms.
2024-10-14
NeurIPS.cc/2024/Workshop/AdvML-Frontiers (publié)
openreview.net
openreview.net
Active Semantic Mapping and Pose Graph Spectral Analysis for Robot Exploration
Rongge Zhang
Haechan Mark Bong
Giovanni Beltrame
2024-10-13
2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (publié)
doi.org
arxiv.org
Local Linearity is All You Need (in Data-Driven Teleoperation)
Michael Przystupa
Gauthier Gidel
Matthew E. Taylor
Martin Jagersand
Justus Piater
Samuele Tosatto
One of the critical aspects of assistive robotics is to provide a control system of a high-dimensional robot from a low-dimensional user inp… (voir plus)ut (i.e. a 2D joystick). Data-driven teleoperation seeks to provide an intuitive user interface called an action map to map the low dimensional input to robot velocities from human demonstrations. Action maps are machine learning models trained on robotic demonstration data to map user input directly to desired movements as opposed to aspects of robot pose ("move to cup or pour content" vs. "move along x- or y-axis"). Many works have investigated nonlinear action maps with multi-layer perceptrons, but recent work suggests that local-linear neural approximations provide better control of the system. However, local linear models assume actions exist on a linear subspace and may not capture nuanced motions in training data. In this work, we hypothesize that local-linear neural networks are effective because they make the action map odd w.r.t. the user input, enhancing the intuitiveness of the controller. Based on this assumption, we propose two nonlinear means of encoding odd behavior that do not constrain the action map to a local linear function. However, our analysis reveals that these models effectively behave like local linear models for relevant mappings between user joysticks and robot movements. We support this claim in simulation, and show on a realworld use case that there is no statistical benefit of using non-linear maps, according to the users experience. These negative results suggest that further investigation into model architectures beyond local linear models may offer diminishing returns for improving user experience in data-driven teleoperation systems.
2024-10-13
2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (publié)
doi.org
PhotoBot: Reference-Guided Interactive Photography via Natural Language
Oliver Limoyo
Jimmy Li
Dmitriy Rivkin
Jonathan Kelly
Gregory Dudek
We introduce PhotoBot, a framework for fully automated photo acquisition based on an interplay between high-level human language guidance an… (voir plus)d a robot photographer. We propose to communicate photography suggestions to the user via reference images that are selected from a curated gallery. We leverage a visual language model (VLM) and an object detector to characterize the reference images via textual descriptions and then use a large language model (LLM) to retrieve relevant reference images based on a user's language query through text-based reasoning. To correspond the reference image and the observed scene, we exploit pre-trained features from a vision transformer capable of capturing semantic similarity across marked appearance variations. Using these features, we compute pose adjustments for an RGB-D camera by solving a perspective-n-point (PnP) problem. We demonstrate our approach using a manipulator equipped with a wrist camera. Our user studies show that photos taken by PhotoBot are often more aesthetically pleasing than those taken by users themselves, as measured by human feedback. We also show that PhotoBot can generalize to other reference sources such as paintings.
2024-10-13
2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (publié)
doi.org
arxiv.org
The Canadian VirusSeq Data Portal and Duotang: open resources for SARS-CoV-2 viral sequences and genomic epidemiology
Erin E. Gill
Baofeng Jia
Carmen Lia Murall
Raphaël Poujol
Muhammad Zohaib Anwar
Nithu Sara John
Justin Richardsson
Ashley Hobb
Abayomi S. Olabode
Alexandru Lepsa
Ana T. Duggan
Andrea D. Tyler
Arnaud N'Guessan
Atul Kachru
Brandon Chan
Catherine Yoshida
Christina K. Yung
David Bujold
Dusan Andric
Edmund Su … (voir 47 de plus)
Emma J. Griffiths
Gary Van Domselaar
Gordon W. Jolly
Heather K. E. Ward
Henrich Feher
Jared Baker
Jared T. Simpson
Jaser Uddin
Jiannis Ragoussis
Jon Eubank
Jörg H. Fritz
José Héctor Gálvez
Karen Fang
Kim Cullion
Leonardo Rivera
Linda Xiang
Matthew A. Croxen
Mitchell Shiell
Natalie Prystajecky
Pierre-Olivier Quirion
Rosita Bajari
Samantha Rich
Samira Mubareka
Sandrine Moreira
Scott Cain
Steven G. Sutcliffe
Susanne A. Kraemer
Yelizar Alturmessov
Yann Joly
Marc Fiume
Terrance P. Snutch
Cindy Bell
Catalina Lopez-Correa
Julie G. Hussin
Jeffrey B. Joy
Caroline Colijn
Paul M. K. Gordon
William W. L. Hsiao
Art F. Y. Poon
Natalie C. Knox
Mélanie Courtot
Lincoln Stein
Sarah P. Otto
Guillaume Bourque
B. Jesse Shapiro
Fiona S. L. Brinkman
Fiona S. L. Brinkman
The COVID-19 pandemic led to a large global effort to sequence SARS-CoV-2 genomes from patient samples to track viral evolution and inform t… (voir plus)he public health response. Millions of SARS-CoV-2 genome sequences have been deposited in global public repositories. The Canadian COVID-19 Genomics Network (CanCOGeN – VirusSeq), a consortium tasked with coordinating expanded sequencing of SARS-CoV-2 genomes across Canada early in the pandemic, created the Canadian VirusSeq Data Portal, with associated data pipelines and procedures, to support these efforts. The goal of VirusSeq was to allow open access to Canadian SARS-CoV-2 genomic sequences and enhanced, standardized contextual data that were unavailable in other repositories and that meet FAIR standards (Findable, Accessible, Interoperable and Reusable). In addition, the portal data submission pipeline contains data quality checking procedures and appropriate acknowledgement of data generators that encourages collaboration. From inception to execution, the portal was developed with a conscientious focus on strong data governance principles and practices. Extensive efforts ensured a commitment to Canadian privacy laws, data security standards, and organizational processes. This portal has been coupled with other resources, such as Viral AI, and was further leveraged by the Coronavirus Variants Rapid Response Network (CoVaRR-Net) to produce a suite of continually updated analytical tools and notebooks. Here we highlight this portal (https://virusseq-dataportal.ca/), including its contextual data not available elsewhere, and the Duotang (https://covarr-net.github.io/duotang/duotang.html), a web platform that presents key genomic epidemiology and modelling analyses on circulating and emerging SARS-CoV-2 variants in Canada. Duotang presents dynamic changes in variant composition of SARS-CoV-2 in Canada and by province, estimates variant growth, and displays complementary interactive visualizations, with a text overview of the current situation. The VirusSeq Data Portal and Duotang resources, alongside additional analyses and resources computed from the portal (COVID-MVP, CoVizu), are all open source and freely available. Together, they provide an updated picture of SARS-CoV-2 evolution to spur scientific discussions, inform public discourse, and support communication with and within public health authorities. They also serve as a framework for other jurisdictions interested in open, collaborative sequence data sharing and analyses.
2024-10-13
Microbial Genomics (publié)
doi.org
Working Backwards: Learning to Place by Picking
Oliver Limoyo
Abhisek Konar
Trevor Ablett
Jonathan Kelly
Francois Hogan
Gregory Dudek
2024-10-13
2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (publié)
doi.org
arxiv.org