Filters








4,583 Hits in 12.7 sec

Avalanche: an End-to-End Library for Continual Learning

Vincenzo Lomonaco, Lorenzo Pellegrini, Andrea Cossu, Antonio Carta, Gabriele Graffieti, Tyler L. Hayes, Matthias De Lange, Marc Masana, Jary Pomponi, Gido M. van de Ven, Martin Mundt, Qi She (+16 others)
2021 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
In this work, we propose Avalanche, an open-source end-to-end library for continual learning research based on PyTorch.  ...  Avalanche is designed to provide a shared and collaborative codebase for fast prototyping, training, and reproducible evaluation of continual learning algorithms.  ...  We would like to thank all its members for the valuable contributions and feedback that significantly improved the quality of the manuscript and the software library.  ... 
doi:10.1109/cvprw53098.2021.00399 fatcat:kdtmclb2qbdp3drxeluki3vazy

Avalanche: an End-to-End Library for Continual Learning [article]

Vincenzo Lomonaco, Lorenzo Pellegrini, Andrea Cossu, Antonio Carta, Gabriele Graffieti, Tyler L. Hayes, Matthias De Lange, Marc Masana, Jary Pomponi, Gido van de Ven, Martin Mundt, Qi She (+16 others)
2021 arXiv   pre-print
In this work, we propose Avalanche, an open-source end-to-end library for continual learning research based on PyTorch.  ...  Avalanche is designed to provide a shared and collaborative codebase for fast prototyping, training, and reproducible evaluation of continual learning algorithms.  ...  In this work, we propose Avalanche, an open-source (MIT licensed) end-to-end library for continual learning based on PyTorch [39] , devised to provide a shared and collaborative codebase for fast prototyping  ... 
arXiv:2104.00405v1 fatcat:kulpm5kojndxdlo3k3b6e5yxra

Avalanche RL: a Continual Reinforcement Learning Library [article]

Nicolò Lucchesi, Antonio Carta, Vincenzo Lomonaco, Davide Bacciu
2022 arXiv   pre-print
In this paper, we describe Avalanche RL, a library for Continual Reinforcement Learning which allows to easily train agents on a continuous stream of tasks.  ...  Continual Reinforcement Learning (CRL) is a challenging setting where an agent learns to interact with an environment that is constantly changing over time (the stream of experiences).  ...  Finally, we are expecting to merge Avalanche RL into Avalanche, striving to provide a single end-to-end framework for all continual learning applications.  ... 
arXiv:2202.13657v2 fatcat:zlhnurhg4ndrhel4taqkcezfay

Agency in an AI Avalanche: Education for Citizen Empowerment

Harry C. Boyte, Marie-Louise Ström
2020 Eidos  
emphasis on altruism in service learning remains an impediment to real empowerment.  ...  They had to join forces to build forts, roads, and libraries. They also established the first public schools.  ... 
doi:10.14394/eidos.jpc.2020.0023 fatcat:mojkwuxwmvbblctxxx6bimk4vi

Modelling risk-taking behaviour of avalanche accident victims [article]

Robin Couret and Carole Adam and Martial Mermillod
2020 arXiv   pre-print
Each year, over 15000 requests for mountain rescue are counted in France. Avalanche accidents represent 39\% of reports, and are therefore our focus in this study.  ...  The availability heuristic is operationalized by three situations, an avalanche accident video, a backcountry skiing video and a neutral context.  ...  If the participant enters the slope, he continues the simulation, otherwise it ends. 2. If he does not give up, a new scene is displayed associated with a choice of trajectory.  ... 
arXiv:2009.09080v1 fatcat:u5vh6azdnbbahfb6w5ztib4sre

Backscatter Characteristics of Snow Avalanches for Mapping with Local Resolution Weighting

Cedric Mount Tompkin, Silvan Leinss
2021 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
With a processing pipeline for avalanche segmentation using a fixed threshold on the backscatter difference we obtain a higher F1 score (0.75) with LRW compared to an unweighted orbit average (F1 = 0.68  ...  based on S1 imagery of an extreme avalanche event on January 4, 2018 in the Swiss Alps.  ...  It still has room for improvements and research such as [15] which focuses on the optimal application of deep learning independently of the preprocessing needs to be continued.  ... 
doi:10.1109/jstars.2021.3074418 fatcat:hhwkf2lbtjgqpbwt6xab6lvuie

An analysis and a comparative study of cryptographic algorithms used on the Internet of Things (IoT) based on avalanche effect

K.D. Muthavhine, M. Sumbwanyambe
2018 2018 International Conference on Information and Communications Technology (ICOIACT)  
Refer to appendix 2 for an overview of Yang's string.  ...  The reason being that it is difficult to learn than other programming languages. Therefore, it is difficult to crack an algorithm written in C++ compared to others.  ... 
doi:10.1109/icoiact.2018.8350759 fatcat:l4k7bvz6pba2jn3camkkptsjv4

Application of physical snowpack models in support of operational avalanche hazard forecasting: A status report on current implementations and prospects for the future

Samuel Morin, Simon Horton, Frank Techel, Mathias Bavay, Cécile Coléou, Charles Fierz, Andreas Gobiet, Pascal Hagenmuller, Matthieu Lafaysse, Matjaž Ližar, Christoph Mitterer, Fabiano Monti (+5 others)
2019 Cold Regions Science and Technology  
By summarizing currently implemented modelling tools specifically designed for avalanche forecasting, we intend to diminish and contribute to bridging this gap.  ...  Lessons learned from currently used methods are explored and provided, as well as prospects for the future, including a list of the most critical issues to be addressed.  ...  Acknowledgements This study was initiated at a meeting during the Breckenridge 2016 International Snow Science Workshop and evolved into this manuscript, thanks to multiple interactions with the avalanche  ... 
doi:10.1016/j.coldregions.2019.102910 fatcat:lwgjqnhembfu5ft63mlm7fleku

Continual-Learning-as-a-Service (CLaaS): On-Demand Efficient Adaptation of Predictive Models [article]

Rudy Semola, Vincenzo Lomonaco, Davide Bacciu
2022 arXiv   pre-print
It provides support for model updating and validation tools for data scientists without an on-premise solution and in an efficient, stateful and easy-to-use manner.  ...  The two main future trends for companies that want to build machine learning-based applications and systems are real-time inference and continual updating.  ...  Avalanche, End-to-End Library for Continual Learning In [13] it is proposed Avalanche, an open-source endto-end library for continual learning research based on PyTorch.  ... 
arXiv:2206.06957v2 fatcat:dbs7fmql6ffjppept27knysii4

Subject Classifying and Indexing of Libraries and Literature (Book Review)

Lea M. Bohnert
1960 College and Research Libraries  
Goliath and the Council Third Annual Report for the Period Ending June 30, 1959, Council on Library Re- sources, Inc. Washington, D. C.: The Council, 1959, 62p.  ...  The Council, in carrying out its charge, seeks to maintain an over-all view.  ... 
doi:10.5860/crl_02_04_323 fatcat:y6cxsw6ncngx3fp7fccz6fdnru

A year of learning for your lifelong leadership journey: Climbing Mount Everest

Steven J. Bell
2012 College & research libraries news  
It requires little effort to inundate one's self with an avalanche of leadership literature, from the Harvard Business Review to Leadership Digital Daily (a seven-day a week newsletter that links to dozens  ...  The great, never-ending debate in the leadership literature is whether it is an innate quality or a learnable skill.  ... 
doi:10.5860/crln.73.8.8809 fatcat:3q74cb4bzvg5ngetzchoqnwuc4

AN AVALANCHE IS COMING ESSAY AN AVALANCHE IS COMING Higher education and the revolution ahead

Michael Barber, Katelyn Saad, Rizvi Foreword, Lawrence Summers, Michael Barber, Katelyn Donnelly, Saad Rizvi
2013 unpublished
The avalanche metaphor is appropriate because the one certainty for anyone in the path of an avalanche is that standing still is not an option.  ...  The fundamental question in An Avalanche is Coming is whether a university education is a good preparation for working life and citizenship in the 21st century or, more precisely, whether it will continue  ... 
fatcat:eo2bsovjcrg67ikucurqqc6pki

XSLT for tailored access to a digtal video library

Michael G. Christel, Bryan Maher, Andrew Begun
2001 Proceedings of the first ACM/IEEE-CS joint conference on Digital libraries - JCDL '01  
This paper begins with an introduction to a few of these interfaces and their implementation history.  ...  The paper concludes with a discussion of next steps planned for the Informedia library work. (Contains 19 references.)  ...  For example, work continues to identify faces within the video library, and name those faces with proper names.  ... 
doi:10.1145/379437.379686 dblp:conf/jcdl/ChristelMB01 fatcat:7ghc3z7hfnbgthk6z56rmwdf5y

Machine Learning of Spatiotemporal Bursting Behavior in Developing Neural Networks

Jewel YunHsuan Lee, Michael Stiber, Dong Si
2018 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)  
This communication presents an application of machine learning techniques to bridge the gap between microscopic and macroscopic behaviors and identify the small-scale activity that leads to large-scale  ...  behavior, reducing data complexity to a level that can be amenable to further analysis.  ...  Regression analysis was performed using scikit-learn, a Python machine learning library. All model trainings were done on a 2.4GHz Intel Xeon E5-2620v3 system.  ... 
doi:10.1109/embc.2018.8512358 pmid:30440408 fatcat:7bypllivwrfcxieoasu5dt2yii

Sequoia: A Software Framework to Unify Continual Learning Research [article]

Fabrice Normandin, Florian Golemo, Oleksiy Ostapenko, Pau Rodriguez, Matthew D Riemer, Julio Hurtado, Khimya Khetarpal, Ryan Lindeborg, Lucas Cecchi, Timothée Lesort, Laurent Charlin, Irina Rish (+1 others)
2022 arXiv   pre-print
The field of Continual Learning (CL) seeks to develop algorithms that accumulate knowledge and skills over time through interaction with non-stationary environments.  ...  Sequoia also includes a growing suite of methods which are easy to extend and customize, in addition to more specialized methods from external libraries.  ...  MonsterKong is pixel-based, lightweight and has an easily-customizable domain, making it a good choice for evaluating continual learning agents.  ... 
arXiv:2108.01005v3 fatcat:s7z3xiszabeo3gldo3t4owxteq
« Previous Showing results 1 — 15 out of 4,583 results